Mask disinfection using atmospheric pressure cold plasma

Objectives Mask usage has increased over the last few years due to the COVID-19 pandemic, resulting in a mask shortage. Furthermore, their prolonged use causes skin problems related to bacterial overgrowth. To overcome these problems, atmospheric pressure cold plasma was studied as an alternative technology for mask disinfection. Methods Different microorganisms (Pseudomonas aeruginosa, Escherichia coli, Staphylococcus spp.), different gases (nitrogen, argon, and air), plasma power (90-300 W), and treatment times (45 seconds to 5 minutes) were tested. Results The best atmospheric pressure cold plasma treatment was the one generated by nitrogen gas at 300 W and 1.5 minutes. Testing of breathing and filtering performance and microscopic and visual analysis after one and five plasma treatment cycles, highlighted that these treatments did not affect the morphology or functional capacity of the masks. Conclusion Considering the above, we strongly believe that atmospheric pressure cold plasma could be an inexpensive, eco-friendly, and sustainable mask disinfection technology enabling their reusability and solving mask shortage.


Introduction
Masks that cover the nose and mouth are one of the main barriers to avoid the acquisition and transmission of microbial diseases, reducing viral transmission by 95% ( Catching et al. , 2021 ;Feng et al. , 2020 ;Ng et al. , 2020 ;Suess et al. , 2012 ). During the last few years, the use of masks has increased due to the COVID-19 pandemic ( Feng et al. , 2020 ). However, 93.64% of individuals hadskin reactions, like itching, stinging, dryness, and acne after mask exposure ( Choi et al. , 2021 ;Damiani et al. , 2021 ). Maskne (mask + acne) is considered as the skin disease that provokes the highest number of dermatologist visits, and it could be aggravated ( Dharmaraj et al. , 2021 ;Sangkham, 2020 ). In short, mask disinfection promotes environmental sustainability.
Other researchers have also pointed out the functional degradation of masks and the presence of residues after ethanol, UV irradiation, blenching, microwave heating, or VHP treatments ( Battelle, 2016 ;Bergman et al. , 2010 ;Smith et al. , 2020 ;Viscusi et al. , 2009 ).
Low-temperature plasmas have been used for mask disinfection, with hydrogen peroxide gas plasma technology being the most widely used. Most authors have used the Sterrad 100NX plasma generator, which requires a vacuum atmosphere and long treatment times ( Ibáñez-Cervantes et al. , 2020 ;Kumar et al. , 2020 ;Viscusi et al. , 2009 ;Wigginton et al. , 2021 ). Viscusi et al. (2009) analyzed mask deterioration after a 55 minutes of hydrogen peroxide gas plasma treatment and observed no negative impact either on filter aerosol or on filter airflow resistance. In some studies, more than one cycle of hydrogen peroxide gas plasma treatment affected the filtration and breathing capacity of masks ( Bergman et al. , 2010 ;Kumar et al. , 2020 ;Wigginton et al. , 2021 ).
In this regard, atmospheric pressure cold plasma (APCP) technology has been a focus of increased research in recent years for disinfection applications. The most important investigations on the use of APCP for mask disinfection have been examined to further elaborate on the novelty of the current study, which, in conclusion, is scarce in the literature. Indeed, to our knowledge, there are only two publications. In the first publication, a dielectric barrier discharge plasma generator was used to generate ozone to treat KF94 mask portions (30 × 35 mm size) at different durations (10-300 seconds) against virus and bacteria . The authors observed a total inactivation of S. aureus (10 seconds). Neither the structural characteristics of the filter layer nor the functional properties of the mask altered negatively after five 60-second treatment cycles in KF94 masks. In the second publication, plasma was generated using a dielectric barrier discharge plasma equipment to treat SARS-CoV-2 inoculated on N95/ filtering face piece 2 (FFP2) and N99/ filtering face piece 3 (FFP3) mask portions (4 × 4 cm size). No significant negative effect was detected on filtration efficiencies. Plasma was characterized by analyzing the optical emission spectrum, and it was observed that reactive oxygen and nitrogen species (RONS) such as OH * , NO, and ozone had a viricidal effect .
In this study, an APCP equipment known as atmospheric pressure plasma jet was used. Different combinations of plasma power, plasma gas, and treatment time were tested on the inner surface of KN95 and FFP2 masks to disinfect them. E. coli, Pseudomonas aeruginosa, and different Staphylococcus spp. were evaluated.
The original contribution of this work was the determination of the optimal plasma treatment parameters to achieve a total inactivation of all bacteria while preserving the physical and morphological properties of the masks.
For this purpose, the antimicrobial activity of both plasmatreated and thermally treated inoculated samples was analyzed. In addition, plasma-treated samples were analyzed by scanning electron microscopy (SEM), visual evaluation, as well as filtration and breathing tests.

Plasma and thermal treatments
Different plasma treatments were applied to sets of three sample masks ( Table 1). The APCP equipment used was PlasmaSpot500 (MPG, Luxemburg) for mask disks and complete mask treatments ( Sainz-García et al. , 2021 ). It consists of one external electrode connected to a high voltage source, one internal grounded electrode, and an aluminum oxide (Al 2 O 3 ) dielectric tube between them. Figures 2 and 3 illustrate the APCP equipment and how mask disks and complete masks were treated, respectively.
The effect of the heat flow generated by the plasma was studied by controlling the temperature during the plasma treatment and by applying thermal treatments in complete masks. The outer surface of the mask (layer 1) subjected to the best plasma treatment (nitrogen plasma 2 [N2]) was characterized by thermography using a thermal imaging camera (TESTO 871) (Testo SE & Co. KGaA, Germany). The obtained images were analyzed with the IRSoft (version 4.3) software. The temperature of the inner surface of the mask (layer 5) was monitored during each second of the 1.5 minutes of plasma treatment by a K-type Teflon-coated thermocouple connected to a data logger (Testo 167T4) ( Figure 3c ). This study was performed to determine the maximum temperature reached for each treatment. Heated gas thermal treatments were applied on inoculated complete masks to analyze the antimicrobial activity of the heat flow ( Figure 3d ).
Three bacteria were inoculated per disk for each treatment as shown in Figure 2 . Three inoculated disks were used for each plasma treatment, including the CT treatment (Table 1) to control bacterial growth (positive control, CT-positive). In addition,  untreated disks without bacteria were used to analyze the presence of possible contaminants (negative control, CT-negative). No colonies were seen in the negative controls regardless of the treatment studied. Plasma treatments were applied after the inoculum was dry. Then, each inoculated area was independently washed in Eppendorf tubes with 300 μl of Mueller-Hinton broth (Con-daLab) for 24 hours. After that, serial dilutions were performed and cultured in Brain Heart Infusion agar at 37 °C overnight to determine the CFU/ml. Furthermore, the antimicrobial effect of the APCP treatments N2 and N3 was also analyzed by quantifying the fluorescence levels of the green fluorescent protein (GFP) in the P. aeruginosa strain ATCC15692GFP. For this purpose, mask disks were inoculated with 10 μl of P. aeruginosa ATCC15692GFP, dried, and then treated with APCP. GFP was quantified before and after the treatment by digital image analysis on a fluorescence microscope (Nikon) using the ImageJ software.

Antimicrobial activity of plasma treatments on complete masks
Five points were selected on the inner side of the masks (Nose-N; Mouth-M; Chin-CH, and both Cheeks-C1 and C2) as shown in Figure 3a . Each spot was inoculated with 10 μl of 0.5 McFar- land suspension of E. coli ATCC25922, S. aureus ATCC29213, and P. aeruginosa PAO1. Three bacteria were inoculated per mask for each treatment. Untreated masks were used as positive controls for growth. Plasma treatment was applied after the inoculum was dried under sterile conditions. Subsequently, each area was cut and the viable bacteria CFU/ml were determined using the procedure described for mask disks.

Optical emission spectroscopy
Optical emission spectroscopy analysis was used to identify the RONS produced by N2, Ar1, and A2 plasma treatments (Table 1). The equipment used was the spectrometer F600-UVVIS-SR (Stellar-Net, Tampa, Florida, USA), connected to an optical fiber (QP600-2SR-Ocean Optics) with a lens that collected information of the plasma flux as shown in Figure 2b . Data were processed with the SpectraWiz software (StellarNet, Tampa, Florida, USA).

Physical and morphological characteristics of the mask after treatment
Filtration properties and breathing resistance of complete masks were measured by an external laboratory (AITEX, Spain). The filtration test was done by filter penetration with paraffin oil following the EN 149:20 01 + A1:20 09 standard. The expanded uncertainty was ± 10% of the measured value for a 95% probability of coverage. The breathing resistance test was done following the same standard. In this case, the expanded uncertainty was ± 18% of the measured value for 95% probability of coverage. Each analysis was performed five times.
For morphological analysis, a HITACHI S-2400 SEM was used. Untreated and 5-time-plasma-treated (N2) (Table 1) mask disks from layers 4 and 5 were tested. Before SEM examination, mask disks were coated with a thin layer of gold and palladium, using a sputtering apparatus to make them conductive. In addition, visual analysis was performed to detect any alterations in the treated masks.

Inactivation on mask disks by plasma treatments
The degree of bacterial inactivation was influenced by the bacterial species analyzed and the modification of the plasma treatment parameters. Figure 4 shows the antimicrobial activity of the plasma treatments against E. coli ATCC25922, P. aeruginosa PAO1, and S. aureus ATCC29213 on mask disks in comparison with the CT-positive.
The best treatments were the nitrogen plasma treatments (N2 and N3) as they achieved total inactivation regardless of the target bacteria. Furthermore, all studied treatments achieved complete inactivation of E. coli and P. aeruginosa , with S. aureus being the most resistant species, consistent with other studies ( Han et al. , 2016 ;Huang et al. , 2020 ;Kayes et al. , 2007 ;Lunov et al. , 2016 ). For instance, Huang et al. (2020) and Han et al. (2016) found higher inactivation values in Salmonella Typhimurium and E. coli , respectively, than in S. aureus after plasma treatment of 96-well microtiter plates and Petri dishes, respectively ( Han et al. , 2016 ;Huang et al. , 2020 ).
Several authors consider the morphological characteristics of bacteria as one of the factors responsible for the differences in bacterial inactivation ( Huang et al. , 2020 ). Both Gram-positive and Gram-negative bacteria possess cell wall peptidoglycans, but Gram-positive bacteria have a thick peptidoglycan wall that allows them to resist plasma damage. Besides, other researchers have identified different action mechanisms depending on Grampositive and Gram-negative bacteria. For Gram-positive bacteria, it has been suggested that RONS play the main role in provoking lipid membrane disturbances after lipid peroxidation of unsaturated fatty acids. Amino acid oxidation results in protein modification, followed by DNA damage and cell death ( Arjunan et al. , 2015 ;Šimon čicová et al. , 2018 ;Yong et al. , 2015 ). However, in the case of Gram-negative bacteria, with irregular surfaces, electrostatic disruption is the most effective effect ( Lunov et al. , 2016 ;Mai-Prochnow et al. , 2016 ). This effect involves the rupture of the membrane when the outer membrane acquires sufficient electric charge. Regarding the bacteria analyzed in our work, E. coli and P. aeruginosa are Gram-negative bacteria, and S. aureus is Grampositive, which could explain the differences in the inactivation rates.
In addition, plasma treatment time, power, or gas, influenced the bacterial inactivation results, as many authors have previously indicated ( Han et al. , 2016 ;Huang et al. , 2020 ;Lunov et al. , 2016 ;Miao and Jierong, 2009 ;Surowsky et al. , 2014 ;Wiegand et al. , 2014 ). Han et al. (2016) demonstrated that the longer the exposure time, the higher the bacterial damage ( Han et al. , 2016 ). This is in accordance with our results as N1 treatment (45 seconds) only achieved 4.96 log (CFU/ml) reductions, whereas N2 (1.5 minutes) or N3 (2.5 minutes), caused total inactivation. Although the treatment time for N4 plasme treatment was the longest (5 minutes), the plasma power was not enough to inactivate all bacteria (220 W). N2 treatment (300 W; 1.5 minutes) resulted in higher inactivation values than N4 (220 W; 5 minutes), indicating that treatment power plays a more important role than treatment time. In this regard, another previous study reported that when the power was increased from 75 W to 125 W, inactivation by plasma treatment against L. monocytogenes, E. coli, and S. Typhimurium also increased ( Kim et al. , 2011 ). Plasma power provides energy to generate RONS. Thus, the higher the power, the larger the amount of RONS generated to inactivate the bacteria ( Laroussi and Leipold, 2004 ;Lu et al. , 2016 ).
Finally, plasma gas played one of the main roles in terms of antimicrobial activity. Comparing the studied gases (nitrogen, air, and argon), the best was nitrogen because it achieved total inactivation regardless of the bacteria used.
The effect of N2 and N3 treatments against P. aeruginosa ATCC15692GFP was also studied by analyzing their antimicrobial activities and the fluorescence levels of GFP ( Figure 5 ). This bacterium possesses a multicopy vector encoding the green fluorescent protein GFPmut3. Both plasma treatments achieved antimicrobial activity ( > 7 logarithmic reductions, data not shown), and they also reduced protein concentration. GFP signals after N2 and N3 treatments were 43% and 54% respectively, in comparison with signal control (100%) with statistically significant differences ( p ≤ 0.05). Hence, it is confirmed that both treatments were effective against P. aeruginosa ATCC15692GFP.

Inactivation of different species of Staphylococcus
Staphylococcus species are one of the most important causes of nasopharyngeal infections and can be implicated in the most aggressive types of maskne ( Daou et al. , 2021 ;Jusuf et al. , 2020 ;Revai et al. , 2008 ;Sun and Chang, 2017 ). For that reason, we investigated different Staphylococcus spp. strains ( S. hominis W220 , S. haemolyticus W1493, S. saprophyticus W1498, S. epidermidis W213, W232 and W1346, and S. aureus W1623 and W1570) from the clinical collection of Molecular Microbiology Area of the Biomedical Research Center of La Rioja (CIBIR), to determine their resistance to plasma treatment. Among them, the methicillin and linezolid-resistant S. epidermidis W213 and W232 , and the methicillin-resistant S. aureus W1570 and W1623 presented multidrug-resistant phenotypes. Tests were performed applying the best plasma treatment identified in section 3.1 (N2) and following the same procedures. All strains, including the multidrug-resistant strains, were completely inactivated after 1.5 minutes of N2 treatment.

Inactivation of the complete mask by plasma and thermal treatments
N2 plasma treatment was chosen to study the antibacterial activity on complete masks as more than six logarithmic reductions were observed against E. coli ATCC25922, S. aureus ATCC29213, and P. aeruginosa PAO1 ( Figure 4 ). In addition, the thermal effect was studied to determine the antimicrobial capacity associated with the flow of heat generated by the plasma.
On the one hand, the thermal images of the outer layer of the mask (layer 1) with the N2 treatment at different times showed a homogeneous thermal distribution ( Figure 6 ), indicating a homogeneous plasma treatment. The temperatures on the outer surface of the mask at the end of the N2 treatment ranged from 80 °C to 90 °C ( Figure 6d ). On the other hand, the maximum temperature of the inner mouth zone of the mask during the N2 plasma treatment was 100 °C ( Figure 7 ). This temperature was used as a guide to heat an 80 standard liter per minute -nitrogen flux using   a thermal system controlled by a temperature regulator, and subsequently, each bacterium was subjected to that nitrogen flux for 1.5 minute. The thermal effect was the main cause of the inactivation of P. aeruginosa and E. coli as both treatments (plasma and heat flux) produced total inactivation of both bacteria ( Table 2 ). However, heated flow was insufficient to inactivate S. aureus, and it could be affirmed that plasma treatment was necessary for the total inactivation of S. aureus in addition to heat flux treatment. To our knowledge, bacteria can be inactivated by a combination of several factors, including time and heat treatment. In this regard, some authors have studied how thermal treatment affects the inactivation of different bacteria ( P. aeruginosa PAO1, S. Typhimurium, Figure 7. Temperature of the inner surface (layer 5) of a mask during plasma treatments (N2, Ar1, and A2) and thermal treatment measured by a type K thermocouple probe.
and E. coli ) in different environments (sardinella meat, snail meat, and cells). They concluded that the higher the temperature and the time, the greater the bacteria reduction ( Gabriel and Alano-Budiao, 2019 ;Gabriel and Ubana, 2007 ;Marcén et al. , 2017 ).
Optical emission spectroscopy Figure 8 shows optical emission spectra (20 0-50 0 nm) of nitrogen, argon, and air plasma. It was possible to identify which relevant RONS appeared for each plasma gas and justify the antimicrobial activity. N2 treatment spectrum ( Figure 8a ) showed different species namely: (i) NO * radical (200-280 nm), (ii) Second positive system (SPS) of nitrogen (296-405 nm), and (iii) First negative system (FNS) of nitrogen (at 394 and 427 nm). SPS and FNS need to be taken into account because of their role in ozone (O 3 ) generation, a species with biocidal capacity and one that does not generate an excited state that emits light ( Girgin Ersoy et al. , 2019 ;Marino et al. , 2018 ;Porto et al. , 2020 ;Wen et al. , 2020 ). O 3 generation is explained as follows .
On the other hand, Ar1 ( Figure 8b ) showed the same species as N2, in addition to the OH * radical (309 nm). Finally, Figure 8c shows the spectrum of A2, where only the nitrogen species SPS and FNS were detected.
In this study, masks were treated using an assembly that conducts the plasma flow to the inner surface of the mask. Therefore, the combined antimicrobial effect of positive and negative ions, RONS, electrons, excited and neutral ions, heat, molecules, and UV photons can be leveraged ( Scholtz et al. , 2015 ). Most authors propose that the effect of RONS is most likely responsible for the inactivation capacity of atmospheric pressure plasma jet ( Iuchi et al. , 2018 ;Sainz-García et al. , 2021 ). In fact, it has been previously reported that the NO * , OH * , and O 3 species confirmed antimicrobial activity ( Kaushik et al. , 2018 ;Laroussi and Leipold, 2004 ;Porto et al. , 2020 ;Wen et al. , 2020 ;Wu et al. , 2017 ;Zhang et al. , 2016 ).
The following conclusions might be drawn when comparing treatments: (i) N2 treatment achieved the best bacterial inactivation ( Figure 4 ), and among all the reactive species identified, the NO * radical could be the most biocidal RONS, as it has the highest peak and intensity ( Figure 8a ). (ii) The antimicrobial capacity of Ar1 was similar to that of A2 despite generating a significantly lower heat flow than the other treatments ( Figure 7 ). In this case, OH * radicals and O 3 from nitrogen species SPS and FNS seem to balance out the lower Ar1 inactivation effect due to the low temperature. (iii) Notwithstanding the lower relative intensities for all RONS (NO * , OH * , and O 3 ), the inactivation performed by treatment A2 could benefit from heat flux (highest temperature) ( Figure 7 ).

Physical and morphological characteristics of the mask after treatment
Filtration capacity (FC), breathing resistance, visual modifications, and adaptability to the face were assessed after plasma   treatments. Table 3 shows FC and breathing resistance after one and five treatment cycles. Data obtained revealed that plasma treatment did not affect either breathing resistance or FC of masks for at least five treatment cycles. It was observed that the resistance to inhalation (30 l/minutes) increased from 0.387 mbar to 0.420 mbar after one cycle and decreased from 0.466 mbar to 0.410 mbar after five cycles, which were not significant differences. In contrast, in the study of inhalation 95 l/min and expiratory resistance, there was a slight increase in one cycle but a decrease after five cycles. Finally, the number of cycles did not affect the FC which was similar to the control masks. These results seem reasonable, and a slight increase in respiration implies a small decrease in FC. Nevertheless, this reduction could be neglected considering that FC of treated masks during five cycles only lost 3.4% of FC with regard to the control masks (FC = 1.418). Our study demonstrated that treated masks maintained a FC of 1.468, similar to FFP3 (FC = 1). These findings are in line with previous results that applied five cycles (60 seconds) of O 3 plasma treatment on masks without observing modifications in the structural characteristics, functional properties and inhalation resistance of the filter layer . Moreover, previous studies observed no modification in the functional characteristics of masks after one cycle of plasma treatment, suggesting that after one cycle, it is possible to reuse masks in terms of FC and breathing resistance ( Bergman et al. , 2010 ;Osaili et al. , 2020 ;Wigginton et al. , 2021 ). It is worth noting that we succeeded with five plasma treatment cycles, as few studies have evaluated more than one plasma treatment cycle without affecting the functional properties of the masks. Other researchers have studied the effect of more than one cycle of other decontamination technologies on the masks. Wigginton et al. (2021) demonstrated a maintained filtration efficacy and proper fit after 10 cycles of VHP and moist heat ( Wigginton et al. , 2021 ). Bergman et al. (2010) showed no differences in FC with three cycles of UV, ethylene oxide or microwaves ( Bergman et al. , 2010 ). Other authors applied 10 cycles of either ozone or steam and 20 cycles of microwaves without affecting filtration efficiency ( Blanco et al. , 2021 ;Ou et al. , 2020 ;Zulauf et al. , 2020 ). Furthermore, it was observed that the plasma treatment applied in this research had no impact on the mechanical and face comfort capacity as well as visual modifications. There were no differences between control masks and treated masks: no fragmented or stretched elastic trips and no degradation in mask color (data not shown). On the contrary, Viscusi et al. (2009) observed that metallic nosebands were tarnished and not as shiny as in the control masks after a 55-minute plasma treatment.

SEM analysis
Layers 5 and 4 from untreated and plasma-treated samples (five cycles) were analyzed with SEM to determine the possible impact of plasma on surface morphology. SEM images of the samples with a magnification of 150x and 1500x, showed no differences ( Figure 9 ). These results demonstrated no negative impact on the fibers of the treated masks, which is probably related to the fact that their FC was maintained. Therefore, it can be concluded that plasma treatment can be applied without provoking a noticeable damage on mask surfaces.

Conclusions
Due to the COVID-19 pandemic, mask-wearing has increased over the last few years. In this scenario, there has also been a huge increase in facial diseases caused by masks. In addition, there is a shortage of masks due to massive use. In this regard, in this study we investigated the effectiveness of an atmospheric pressure plasma jet equipment in disinfecting masks and maintaining their functional properties. It is worth mentioning that, to the best of our knowledge, this is the first study in which a complete FFP2 mask was disinfected using APCP. Our results suggested that plasma power, plasma gas, and treatment time must be considered to determine the degree of bacterial inactivation. The longer the treatment time and plasma power, the higher the inactivation. Regardless of the type of bacteria, the use of nitrogen plasma, 300 W power, and 1.5 minutes of treatment were the optimal plasma parameters for total inactivation of bacteria. Furthermore, the thermal study confirmed that both P. aeruginosa PAO1 and E. coli ATCC25922 were inactivated mainly by means of the thermal effect. In contrast, RONS generated in plasma were the main cause of Staphylococcus inactivation. Specifically, NO * radical seems to be the most biocidal radical.
FC, breathing resistance analysis and SEM images confirmed that neither reduction in FC nor morphological modifications occurred in the masks even after five cycles of plasma treatment.
In conclusion, disinfection of used masks with APCP could be an emergency solution to reduce facial infections and to solve mask shortages as well as the environmental problems associated with discarding masks. Moreover, this disinfecting technology could be applied to other objects and personal protective equipment used in hospitals.

Funding
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Ethical approval
Not applicable since this was a laboratory study not involving any clinical samples or human and animal subjects .

Data availability statement
Data that support the findings of this study are available from the corresponding author upon reasonable request.

Declaration of competing interest
The authors have no competing interests to declare.