Assessment of N95 respirator for reuse after sterilization: filtration efficacy, breathing resistance, quality factor, chemical structure and surface charge density

In the present ongoing pandemic, the N95 respirator is an essential protective barrier to suppress the spread of the SARS-Cov-2 virus and protect the frontline worker from exposure. The N95 respirators are meant for single usage; however, they can be used after decontamination in-light of the economy and shortfall in availability. At this juncture, the respirators performance after various types of sterilization and usage condition is required to be analyzed in detail. With this motto, this work has proceeded. The filtration efficiency, pressure drop, and quality factor of the respirator are evaluated for two face velocities (5.8 and 26.4 cm/s) following different sterilization methods. Sterilization techniques used here are dry air oven heating, gamma irradiation, and immersing in a 10% concentration of liquid hydrogen peroxide. The particle filtration performance and electrostatic surface charge density measurement are used to determine the facemasks efficacy after sterilization. The methods recommended to sterilize N95 masks without affecting their performance are (i) using dry air heat at 80oC and (ii) H2O2 soaking. The highest reduction in filtration efficiency is observed to be 30-35% after gamma irradiation due to a change in the electrostatic properties of the respirator layers. However, the filtration efficiency does not change significantly for other sterilization methods despite a change in charge density, but there is no direct correlation with filtration efficiency. Electrostatic charge measurement of the filtration layer is a crucial indicator of filtration efficiency degradation. Policymakers can use these data during potential future N95 shortage to assess the viability of sterilization methods.


Introduction
Personal Protective Equipment (PPE), such as respiratory facemasks, offer protection against the spread of COVID-19through droplet and aerosols transmission (Wang and Du, 2020). When an infected person speaks, sneezes and coughs, the virus-containing droplets (> 5.0 µm) and aerosols (< 5.0 µm) will be excreted, remain suspended for up to a few hours and travel a longer distance on the air (Prather et al., 2020 andGuzman, 2020). N95 is essential PPE to protect the wearer with at least more than 95% filtration efficiency for Most Penetrating Particle Size (MPPS) 0.1 -0.3 µm and higher efficacy for those particles smaller or bigger than MPPS (Qian et al., 1998 and Centers for Disease Control and Prevention, 2020). Hence, N95 offers excellent protection when they correctly seal over the face, where filter fabric made of Nylon, cotton, polyester, and polypropylene (Chawla, 2016). Even though the N95 facemask designed for single use, the recent pandemic crises made it necessary to seek emergency alternatives. Toward this crisis, facemasks are being reused multiple times after sterilization with the appropriate method without compromising effectiveness. Various decontamination processes on the physical and chemical change in the fiber of N95 respirator have been studied in the current work. The sterilization methods used in this work are (i) ionizing radiation (15 and 25 kGy), (ii) dry heat (30 and 60 minutes) and (iii) soaking in 10% concentration of hydrogen peroxide for 30 minutes. The respirators first tested for filtration performance, then were sterilized using the methods mentioned above and tested for filtration efficiency. Here, we carefully studied the filtration efficiency and facemasks charge density to detect a relation to the electrostatic filtration before and after various sterilization. In addition to that, the chemical change to the facemask's fiber was examined after various decontamination.
Most of the current works focus nearly exclusively on the respirator's filtration performance . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) analysis, results and discussion, conclusion are described in this article.

Protective respirator
The certified N95 respirators (Magnum) by the National Institute for Occupational Safety and Health (NIOSH) is used in the present studies. The N95 respirators (Make: Magnum Health & Safety Pvt. Ltd., India) is a pouch/ duckbill type, and an optical image is shown in Fig. 1. The Filtration Efficiency (FE) of NIOSH certified respirator is more than 95% for most penetrating

Material test setup
The experimental setup consists of an aerosol's generator, aerosols diagnostic instruments and a differential pressure monitor. The respirators are evaluated in the HEPA Filter Testing Laboratory, Indira Gandhi Centre for Atomic Research (IGCAR). The facility consists of a stainless-steel cylindrical duct which has provision for fixing the facemask without any leakage.
. CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

Sterilization/ decontamination methods
The following sterilization methods are employed viz. ionizing irradiation, dry heat, and soaking in H 2 O 2 liquid in the present study. The respirators were irradiated with ionizing radiation of two doses, 15 and 25 kGy, using gamma chamber GC 500 having a dose rate of 1.80 Gy/h. The typical radiation dose for medical equipment sterilization (24kGy) and deactivation of virus minutes. Special care has been taken in each decontamination process not to bend or damage the respirator to avoid the reduction in filtration efficiency, decreasing surface charge density and mechanical properties. An optical image of the gamma irradiation chamber, hot air oven, and H 2 O 2 sterilization are given in supporting information (Fig.S1). The facemask, which is liquid submersion (soaking in H 2 O 2 ) has been dried overnight before performing the FE testing.
. CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted May 27, 2021.

Data acquirement and analysis
The FE of the respirator is calculated before and after decontamination using the following formula: where C up and C down are the average aerosols concentration in counts/liter of upstream and downstream. The sampled data has been recorded for about two three minutes in 6 seconds frequency of each respirator from aerosols spectrometer (total sample record 20 30). FE estimated by taking the average of the collected data (upstream and downstream) for each mask tested period. FE averaged for all similar type masks and tested conditions (face velocity and sterilization methods). The uncertainty associated with FE of respirator has been calculated based on Type A and Type B uncertainty (Kumar et al., 2020 b). After that, the combined uncertainty has been assessed and then expanded uncertainty resulting from combined uncertainty by multiplying the coverage factor (95% confidence level).

Surface charge measurement
The electrostatic surface charge density is determined by measuring the surface charge of the respirator. The Keithley electrometer (Make: M/s. Tektronix USA and Model: 6517B) used for charge measurement of the respirators, and the schematic presentation is given in Fig.2.The electrometers input is a three-lug (Triax) connector, with the red-colored innermost wire (input high) is the charge-sensing terminal. As shown in Fig. 2, the charge-sensing terminal is connected to the metal surface (Copperplate). The metal noise shield (another copper enclosure) connected with black wire (input low), and the third wire (green color) is grounded. Both copper plate and noise shield are electrically isolated by introducing an insulating Teflon surface in . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted May 27, 2021. ; https://doi.org/10.1101/2021.05.25.21257801 doi: medRxiv preprint 7 between. The charge measuring range by Keithley electrometer is 0.002 to 2 µC with an accuracy of 0.4% of the measured value. The test sample is a piece that has an area of 10 cm 2 (4 cm x 2.5 cm) cut from separated each mask layer. The surface charge on each layer for the respirator has been measured and recorded.

Differential pressure drop across the facemasks
The measured pressure drops (ΔP) across the respirators before and after sterilization is given in  . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

Evaluation of filtration efficiency of masks for aerosol optical size > 0.3µm
FE of control facemasks was determined for the particle of optical diameter >0.3 µm, and two face velocities ( Table 1). As expected, the FE of certified N95 masks is greater than 95%, and FE is more for low face velocity and less for high face velocity. The mean FE of control masks is 99.73±0.08 for low face velocity and 95.31±0.50for for high face velocity. The FE of respirators remained the same after dry heat and H 2 O 2 sterilization for low face velocity. The FE slightly increased after dry heat for high face velocity and remained the same after H 2 O 2 sterilization.
The FE significantly decreased after gamma decontamination for both flow conditions. The certified face masks (N95) have an electret fibers filter that is efficient at low airflow velocity rather than high airflow velocity (Colbeck and Lazaridis, 2014).

Quality Factor (Q f )
A commonly used relation for filtration quality factor to determine the performance of the respirator defined as (Huang et al., 2013, Lin et al., 2017 a, b): Where FE is in % and Δ P is in Pa. A higher FE and lower Δ P yield a higher filtration quality factor, i.e., better performance of the respirators. When the FE is ≥ 95%, the quality factor is more than0.031 Pa -1 and 0.022 Pa -1 for inhalations and exhalation resistance limit at 345 and 245 Pa pressure drop. The quality factor of the facemask for both face velocity is included in Table 1. However, it is found less than the defined limits for high face velocity due to high-pressure drop across the mask (except gamma sterilization). In the case of ionizing radiation decontamination, the Q f value is less due to FE reduction.

Structure component of respirator
The after various sterilization processes. However, after H 2 O 2 sterilization, the printed matter on the surface of facemasks is appeared to be faded.

Chemical change of respirators fiber
The N95 respirator comes in many brands and models that use different materials and method to . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

Fig.3 A typical ATR-FTIR spectra in the transmission mode of the filter layer of control and decontaminated respirators.
The ATR-FTIR spectrum of respirator also contains two intense peaks and various small peaks in 1500 to 750 cm -1 . The peak at 1457 cm -1 is attributed to CH 3 asymmetric deformation or CH 2 scissor vibrations, while the peak at 1373 cm -1 is due to CH 3 symmetric deformation vibrations  (Fig 3). It can be concluded that gamma irradiation and other sterilization methods have not caused any significant changes in the molecular structure of PPs.
. CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted May 27, 2021

Electrostatic surface charge density (σ)
The electrostatic charge of respirators is measured before and after decontamination to evaluate the effect of surface charge density on FE. The average surface charge of a full respirator (without cutting and separating each layer) is measured and summarized in table S1. The average surface charge on the control respirator is + 0.841±0.441, which is in line with other literature (Hossain et al., 2020). The polarity of the surface charge on the respirator changed after sterilization; however, the average charge magnitude has increased despite decontamination methods. The variation in the average surface charge of respirators from a control mask ranges from 351-992%. The highest variation observed after gamma sterilization, while the lowest variation in dry heat for 60 minutes. Here, we observed that the full respirators charge measurement does not convey any correlation with filtration performance. However, FE has decreased after gamma sterilization, but the average charge on the whole surface increased.
To investigate more insightful information on the surface charge on each layer of respirator has been investigated. As explained earlier, the respirator structure consists of three layers. Surface charge measurement is carried out by separating each layer, and care has been taken not to damage the filter media, which would alter the electrostatic charge measurement. According to The charge density imposed on all layers of the respirator has almost the same order. After sterilizing the respirator by dry heat and H 2 O 2 , the polarity of charge density remains similar, but the magnitude persists for dry heat and increases for H 2 O 2 . After gamma irradiation, the charge density on the respirator changed the polarity of the outer and filter layer but remains identical for the inner layer; however, after gamma irradiation, a reduction in charge density is found significantly large. The calculated σ is one order less than the reported values of N95 and K95 (Yim et al., 2020). The measured surface charge density has a significant statistical variation.
The absolute values are uncertain; however, the substantial change in charge density detected qualitatively. The electrostatic charge of control masks showed that the filtration layer has a significant amount of surface charge. It appears that the filtration layer reduced its surface charge . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. Nevertheless, the finding is consistent with other studies that have observed the degradation in the FE of respirator associated with filter media electrostatic degradation.
Further, the filter layer composed of two layers, both layers carefully separated, and surface charge measured on both filter layers of respirators and given in Table S3. One filter layer is negative polarity, and the other is positive for the control respirator. After sterilization of the respirator, the polarity of negative charge layers has not changed, but the charge magnitude has been altered. The positive charge layer polarity also does not change except gamma irradiation (positive polarity change become negative polarity). The average charge variation on the filter layer of respirators from the combined filter layer (both filter layers are not separated) ranges from 15-56% except 25 kGy sterilization (-31%). The negative variation in average charge means the separated filter layer has given more measured surface charge, while positive variation means the measured surface charge less on separated filter layers.

Assessment of facemask effectiveness after sterilization
The percentage change in FE and σ after different sterilization from control facemasks summarizes in table 3. The Δ FE of the mask from control to dry heat and H 2 O 2 sterilization is less than 1% and -3 %for low and high face velocity. However, FE changes around 29 -36% for gamma sterilization for both flow conditions. The Δ FE is more for gamma sterilization and further increases for higher face velocity. The change in surface charge density on the outer filter and an inner layer of masks from control to dry heat sterilization ranges from 3.05 to -20.55%, . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted May 27, 2021 It is known that aerosol filtration occurs by five basic mechanisms viz. gravitational settling, inertial impaction, interception, diffusion and electrostatic force (Hinds, 1999 andVincent, 2007). The N95 mask consists of electrets fibers media capable of capturing and retaining fine particles through electrostatic and mechanical filtration processes (Myers and Arnold, 2003). The least efficiency is associated with particles in the range from 100 500 nm, more prominent for diffusion and lesser for interception; hence, the 95% efficiency achieved for this range of particles by electrostatic force and dielectrophoretic force. When the electret media modified its charges (surface charge density), the efficiency may also get altered. It is found from the present . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) studies that each layer of respirator and filter layer itself has not only different charge density but also the different polarity. The filter layer is composed of opposite charge layers, which gives rise to a strong electric field gradient inside the filter layers. In the absence of a strong inhomogeneous electric field, the respirators filtration properties are reduced, which is observed in the ionizing radiation sterilization method.

Summary and conclusion
The certified N95respirator evaluated for particulate FE, pressure drop, quality factor, and various disinfection methods for two face velocities 5.8 and 26.4 cm/s. As expected, the pressure resistance increases with increasing face velocity; this reflects that it is harder to breathe through a facemask when the respiratory flow is large. The measured pressure drops during masks testing found to be the accepted range for the inhalation and exhalation resistance. The quality factor of facemasks is in the acceptable content except after gamma irradiation. The FE found to be more than 95% for the control respirator and both face velocity. The FE is similar before and after dry heat sterilization for 30 and 60-minutes duration and reduces to about 70% after gamma sterilization for low face velocity and still lesser with higher face velocity. After H 2 O 2 sterilization, the FE of facemasks remains unaffected. Thus, it is recommended that the N95 respirator can be sterilized a few times with dry heat and H 2 O 2 .
N95 masks, which made of electret filtering media, are not recommended for sterilization or decontamination by ionizing radiation; it will compromise the filtering efficiency. The FTIR spectra of facemasks after gamma irradiation and other sterilization have not caused any significant chemical changes, i.e., the molecular bond nature of the studied filter layer of respirators. The most significant reduction in charge density observed after gamma sterilization . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. However, it is worth noting that the lesser sample size and large variability in charge measurement made it challenging to obtain high certainty quantifiable measurements.
Nevertheless, the finding is consistent with other studies that have observed the degradation in FE of masks associated with filter media electrostatic degradation. Further, the N95 respirators come in many brands and models that use different materials and fabrication methods to meet industrial filtration requirements. The electret filter charge measurement can be assessed for other brands of respirators and sterilization methods for their reusability. However, the behavior of charge density variation in the filter layer would be similar.

Acknowledgement
The authors acknowledge Shri S. Athamalingam, Associate Director, Health, Safety and Environmental Group, and Dr. R. Venkatesan, Head, Radiological and Environmental Safety Division for their encouragement and backing to carry out this work.
. CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted May 27, 2021. ; https://doi.org/10.1101/2021.05.25.21257801 doi: medRxiv preprint . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted May 27, 2021. ; https://doi.org/10.1101/2021.05.25.21257801 doi: medRxiv preprint