Abstract
As a major international public health emergency, COVID-19 has posed many challenges for healthcare professionals who have been heavily exposed to contamination. This article describes the development of a high-filtration capacity mask consisting of filter-element layers interspersed with super-activated carbon fiber fabric, non-woven polypropylene for dental–medical–hospital use and antiviral polyamide with nanostructured SiO2 thin film coating. The study found 98.18% particle filtration efficiency and determined 2.11 mmH2O/cm2 differential pressure, while fluid repellency complied with Brazilian standard NBR ABNT 15052:2004.
COVID-19 (coronavirus disease 2019) is caused by the SARS-CoV 2 virus (severe acute respiratory syndrome coronavirus 2), which attacks the airways 1 and may be transmitted by aerosols or airborne respiratory microparticles, larger respiratory droplets or direct contact with contaminated surfaces. 2 On 11 March 2020, the World Health Organization (WHO) declared a pandemic, defined as a “public health emergency of international concern.” 3
Some COVID-19 patients develop mild illness, but others may experience difficulty breathing, pneumonia and even respiratory failure.
As preventive measures for the general public, WHO recommended social distancing and hygiene measures, such as hand washing or sanitizing with 70° GL alcohol. 3 Personal contact was to be avoided regardless of the other individual’s health status, since COVID-19 has a long incubation period and asymptomatic persons may be transmitting the disease. 1 Note that wearing masks is mandatory in some places. Wearing a mask prevents virus transmission through saliva droplets expelled when speaking, sneezing or coughing.
WHO data showed that 8–10% of healthcare workers were affected by COVID-19 and the respiratory tract was the main site of contamination. 4 This scenario shows that healthcare professionals’ exposure to toxic and pathogenic substances is a serious concern that may add to health risks, as Sessink et al.5,6 and Sato et al. 7 , 8 have shown in studies supporting this assertion. The main routes for occupational exposure are inhalation, epicutaneous absorption and ingestion.
Although already commonly used by health workers, masks have become a way of protecting personal and public health during the COVID-19 pandemic. Initially, the covers recommended for the use of the population were disposable surgical masks and face protection pieces. Worldwide demand led to the dissemination of the production of homemade fabric masks. 9 The Centers for Disease Control and Prevention (CDC) and the WHO understand that homemade masks can only be used as a final prophylactic measure against large respiratory droplets 10 and that fabric masks provide individuals with access to personal protective equipment. 9
Other homemade alternatives began to merge to meet the demand for masks and materials for hospital use, such as the construction in three-dimensional (3D) printing of support for the placement of a reduced size filter, thus obtaining an efficient cover with the application of scarce filtering material. 11
This article, therefore, describes the development of a high-filtration capacity mask consisting of layers of filter elements interspersed with several materials, such as super-activated carbon fiber fabric, non-woven polypropylene for mask for dental–medical–hospital use and antiviral polyamide (PA) fabric, subjected to nanostructured coating with thin SiO2 film, thus conferring superhydrophobic properties (impermeable to water).12,13 These masks are mainly intended for use by first responders and health professionals in hospital settings. Concerning innovative aspects, a system of continuous plasma-assisted treatment of fabrics by electric discharges was developed, used mainly for cleaning and sterilization (decontaminating fabric), which precedes the process of coating ceramic filters based on silicon oxide to boost filtration efficiency and surface wettability alteration aspects.
Literature review
Types and specifications of masks used in a hospital setting
There are two types of masks in regulatory stipulations for medical and hospital use: surgical masks and filtering facepieces (FFPs; FFP2 or N95). 14
A surgical mask is a single-use protective equipment item covering the mouth and nose that is recommended for surgical or critical procedures. It should protect patients from contaminating agents coming from a healthcare professional's airways while also protecting healthcare professionals against contaminating agents from patients; it must also prevent the passage of blood and other body fluids.4,14
Surgical masks must be made from non-woven material for dental–medical–hospital use and must have at least one inner layer and one outer layer. A filtration element (consolidated or not) is required per two Brazilian technical standards: “ABNT NBR 15052:2004 – Non-woven articles for dental-medical-hospital use – Surgical masks – Requirements” and “ABNT NBR 14873:2002 – Nonwoven for articles in dental-medical-hospital use – Determination of bacteriological filtration efficiency.”15,16
The non-wovens used are SMS (spunbond–meltblown–spunbond) composite structures; the spunbond comprises the layer of fibers forming the fabric that is then consolidated by mechanical, chemical or thermal processes. The meltblown is a non-woven fabric made from polymers blown in extruders to form a fabric by the combination of interweaving and cohesive adhesion. In the SMS composite structure, the spunbond fabric affords strength and resistance to abrasion, while meltblown forms a liquid and particle barrier. The range of spunbond–meltblown structural combinations available includes SMS, spunbond–meltblown–meltblown–spunbond (SMMS) and spunbond–spunbond–meltblown–meltblown–spunbond (SSMMS), among others, depending on the properties required for the products. 17
A FFP is an individual protection equipment (EPI) item that covers the mouth and nose to provide an adequate seal on the user's face; there is an efficient filter to retain atmospheric contaminants found in workplaces in aerosol form. To afford protection against aerosols containing biological agents in hospital environments, the FFP must be rated FFP2S or higher. Class 2 particulate filter (FFP) respirators (N95s or equivalents) must be partly or totally made of filtration material that supports handling and use throughout the entire period for which it was designed, in order to comply with Brazil’s technical standards (ABNT NBR 13698:2011 – Respiratory protection equipment – Filtering facial piece for particles, and ABNT NBR 13697:2010 – Respiratory protection equipment – Particle Filters).18,19
Activated carbon fiber fabrics
Activated carbon-based materials have traditionally been used to adsorb chemicals due to their very large specific surface area and high micropore volume, hence their widespread use in toxic gas nanofiltration processes, 20 virus-contaminated blood plasma filters, 21 atmospheric pollutant filters 22 and molecular sieves (gas separation).23,24 However, a nanofilter must have the proper pore size distribution for its function. Yokoyama et al.21 used membranes with 30 nm pore distribution to filter viruses from blood plasma. However, researchers have shown that in the presence of certain types of amino acids, viruses may be aggregated and effectively removed by a filter with a pore size larger than the size of the viruses. On the international market, there are many masks containing filter elements based on activated carbon fiber (ACF) felts, activated carbon and advanced carbon materials, such as graphene. 20 These masks are widely used to absorb toxic gases by wearers exposed to highly polluted atmospheric environments. They are also worn by health professionals exposed to environments contaminated with toxic substances, especially in cancer and epidemic treatment settings.7,20
Activated carbon fiber fabric is a textile artefact that aggregates the adsorption power of activated carbon with the innate characteristic of a carbon fiber-based product. ACF is a fibriform activated carbon. ACF is made from cellulosic, polyacrylonitrile (PAN), phenolic and pitch fibers; it consists of an advanced carbonaceous material, unlike the usual granulated or powdered activated carbon. The diameter of ACF is approximately 5–30 μm, which is very fine compared to granular activated carbon (grain diameter 500–5000 μm), and it has a larger specific surface area. Since the ACF pore diameter is approximately 15–20 Å and most of the pores are micropores, its adsorption and desorption rates are much faster.
The main differences between ACFs and activated granular carbon are the high specific surface area, high adsorption capacity and high liquid or gas phase adsorption kinetics. ACF-based filters have a specific surface area around 200 times greater than granular activated carbon. 22
Surface modification of textiles by plasma-assisted activation and deposition processes
Pre-treatment and finishing of textiles using non-thermal plasma technologies are increasingly being preferred as a surface modification technique by incorporating functionalizing nanostructures, since the latter lead to an increase in the area of activity and consequently in activation efficiency.25–27
Non-thermal plasmas are particularly suitable for use in textile processing because most textile materials are heat-sensitive polymers. Furthermore, it is an aggressive technique only in terms of the reactivity of the medium, whereby a wide range of chemically active functional groups may be incorporated into the textile surface. In general, plasma treatment of a polymer produces significant changes in wettability and adhesion due to changes in chemical composition, molecular weight and surface layer morphology.
The plasma polymerization technique with dielectric barrier discharges (DBDs) operating at atmospheric pressure is used in the application of ultrathin and nanostructured films of metallic oxides (SiO2).28,29 Due to surface reactions, the DBD technique allows extensive two-dimensional porous surfaces to be covered with high levels of conformity and uniformity, so new functionalities may be improved/added to the substrate treated with prolonged action. 30
Materials and methods
Mask architecture and materials
The FFP2 filter developed in this study consists of five layers, as shown in Figure 1.

The filter-element architecture developed in this proposal. PA: polyamide; SSMMMS: spunbond–spunbond–meltblown–meltblown–spunbond.
The outer layer in contact with the environment (A) and the final layer in contact with the wearer's face (E) consist of knitted fabric made with 1 × 80/68 PA 6.6 yarn supplied by Rhodia Solvay, with 20 denier elastane, made on a circular loom with 38 needles per inch with a weight of 180 g/m2.
The PA fabric used in this study was made from Amni Virus-Bac Off antiviral yarn.
The first filter element consists of layers of non-woven (B) polypropylene with an analytic certificate issued by Fitesa do Brasil from a bobbin made on 28 March 2020 with 95.96% bacterial filtration efficiency and 3.30 mmH2Ocm−2 breathability per a report issued by Controlbio (analysis report #96757-2020).
After non-woven filter layers, two layers of super-activated and super-hydrophilic carbon fiber fabric (C and D) imported from Chemviron Carbon-Cloth Division were added. Their specific surface area is greater than 1000 m2 g−1 and the surface is predominantly microporous with a surface density of 120 g/m2.
SiO2 films
For the deposition of silicon dioxide (SiO2) on the PA fabric, a silicic acid (Si(OH)4) 13 prepared by ITA’s Plasmas and Processes Lab was used as a precursor. The solution was obtained using aqueous sodium metasilicate (Na2SiO3·5H2O) at 10% by weight and ion exchange resin (IR-120; Rohm and Haas), with a solution concentration of 0.5 mol/L Si.13,31,32
Polyamide 6.6 plasma treatment process
The PA 6.6 fabric was treated by plasma operating at atmospheric pressure and discharge power of 220 W at 56 kHz. As Figure 2 shows, the treatment begins with the air plasma activation process (activation reactor) to make the textile surface hydrophilic (contact angle greater than 120°), increasing and potentiating absorption, so that it can then quickly incorporate the coating precursors (SiO2) nebulized in the deposition reactor. In the process region, a “hybrid corona-DBD” system for discharge at atmospheric pressure operates in continuous treatment of conjugated activation and deposition processes, unlike other systems found in the literature.

Schematic design of the plasma fabric treatment system. All rollers (grounded electrodes) and high-voltage electrodes may have their position adjusted.
Figure 3 shows details of the processing region in the activation and deposition reactors, showing the filamentous character of the DBD-type discharge that acts directly on the fabric surface. Figure 3(a) shows the three discharge regions generated in the three upper high-voltage electrodes, treating the upper side of the fabric that surrounds the ground electrode coated with high-temperature silicone dielectric. Figure 3(b) shows the two lower regions of discharge generated in the high-voltage electrode placed between two dielectric coated grounded rollers, treating the underside of the textile. In the case of the polymerization process, the precursor is inserted through open microchannels in the high-voltage electrodes. The distance between the high-voltage electrode and the sample was set at 2 mm for these processes. This distance was defined experimentally by finding a distance at which the discharge operates stably with good spatial uniformity. A high-voltage source – not more than 40 kV and 60 kHz – was used to supply electrical power for the discharge.

View of the plasma treatment region between the central polarized (high-voltage) electrode and dielectric (colored red): (a) electrode region with three electrodes and three discharge regions over the upper roller, which treats the upper side of the fabric and (b) electrode region with one electrode and two discharge regions between the two lower rollers, which treats the underside of the fabric. (Color online only.)
Determining the contact angle
The effects of plasma treatment on polymeric surfaces must be determined in order to see how the treatment affects surface tension. One of the fabric characteristics that is heavily impacted by plasma treatment is wettability. One of the techniques widely used to evaluate the interaction between a fluid surface and solid or surface wettability is the contact angle. For this study, the contact angle technique was used to assess the effects of plasma on mesh fabric on plasma-treated and untreated samples. Surface properties were evaluated using a goniometer for the sessile drop, which is a direct method of obtaining the contact angle: a drop of distilled water was placed on the surface and the contact angle was obtained using an FTA 1000 device. 28
Chemical characterization of the surface
To identify chemical changes generated by plasma treatment and silica deposition, treated and untreated samples were analyzed using attenuated total reflectance Fourier transform infrared spectroscopy (FTIR-ATR). For this evaluation, Perkin Elmer's Spectrum Spotlight 400 equipment was coupled with ATR accessories. The spectra were obtained by transmission and 64 scans were made to obtain each spectrum at 4 cm–1 resolution. A background spectrum was obtained before obtaining the spectrum, thus enabling the surface layers of the samples to be analyzed by scanning from 450 to 2500 cm−1 in the activation process and from 500 to 4000 cm−1 in the deposition process.
Field emission gun scanning electron microscopy
An electron microscope with a Tescan/Mira 3 brand/model field emission gun was used to assess the microstructural characteristics of activated samples and polymerized coatings.
Particle filtration efficiency, determining differential pressure and fluid repellence
Filters were assembled from several textile layers following the architecture shown in Figure 1. These filters were sent to Brazil’s Technological Research Institute (IPT) to analyze their conformity with ABNT NBR 15052:2004 specifications in terms of particle filtration efficiency (PFE), breathability or pressure differential, and fluid repellence. Once the filters had obtained certification, cutting and sewing services were engaged to make the masks.
Results and discussion
Contact angle analysis
Figure 4 shows the decay curve showing the contact angle varying over time (t).

Contact angle decay curve and ageing time for air plasma-activated fabric. The contact angle of the untreated sample was 120° (so it was hydrophobic).
After applying 220 W air plasma discharge, the sample was placed in a conditioned environment without vacuum preservation in order to measure the decay effect. Zero is the point in time from which the contact angle may be assessed since the surface, immediately after plasma treatment, is super-hydrophilic and the contact angle is zero. Table 1 shows the respective values as of the point in time at which the contact angle may be assessed.
Contact angle decay (220 W)
The contact angle of the untreated sample was 120° (i.e., hydrophobic). The results show that the fabric becomes super-hydrophilic immediately after treatment and that post-plasma exposure time is crucial to altering wettability. In addition, there is a variation in the contact angle, identified in the analysis performed after 24 hours, which relates to polymer mobility, which allows the surface to be restructured hours after plasma treatment. 33
Coating fabric by plasma polymerization
Figure 5 shows the FTIR-ATR spectrum of the PA 6.6 sample before and after deposition of silicon dioxide (SiO2) films using a silicic acid (Si (OH)4) solution as the precursor. The untreated PA (Figure 5 – red line) shows inherent band of nylon at 3290 cm−1 attributed to N-H stretching vibrations. The 2920 and 2850 cm−1 peaks are related to the CH2, asymmetric and symmetrical stretching vibrations respectively, while the 1630 cm−1 absorption band is attributed to the C=O carbonyl stretching vibration of the secondary amide band (amide I).34,35 The 1540 cm−1 amide II band may be attributed to N-H bending motion and the 680 cm−1 band to the O=C−N group.34,36 After plasma treatment (Figure 5 – black line), the significantly higher intensities for N-H and both asymmetrical and symmetrical C-H stretching vibration bands at 3290, 2920 and 2850 cm−1, respectively, may be attributed to the formation of low molecular weight material due to the DBD treatment of PA fabric. 37 According to the literature, ion bombardment (N2+, N+, O2+, H2O+, O2− and O−) induced by plasma discharge may cause bond breakage with energy below 10 eV in the outer layers of the polymer, especially the C-N bonds (the weakest in the polymer chain).34,38 The intensification of the characteristic peaks for −CH3 (1380 cm−1), −CH2 (1470 cm−1) and CH (1420 cm−1) also suggests the formation of hydrocarbon fragments on the surface of the plasma-treated PA fabric. 39 PA fabric has two characteristic crystalline peaks at 930 cm−1 (axial strain of C–C=O amide) and 1200 cm−1 (symmetric out-of-plane angular strain, amide III). Plasma treatment altered peak intensities in these regions, indicating significant changes in crystallinity and C-C bond structures.

Infrared spectrum for the control sample (untreated, red line) and sample coated with SiO2 thin film via plasma (black line). (Color online only.)
In Figure 5, the black line shows the four peaks arising from Si–O stretching vibrations in the 940–1140 cm−1 region, namely at 1116, 1085, 1047 and 1024 cm−1,40–42 are inserted in the peak highlighted by the dashed lines, but Si–O stretching peaks are noticeable at 1024 and 1085 cm−1, which become substantially more intense after the plasma deposition process, leading to a higher nano-silica concentration on the surface. Furthermore, the silanol group (Si-OH) ∼ 3670 cm−1 is one of the key indicators showing that hydrogen species are trapped by the oxide layer that formed after heat treatment.43–45
Figure 6 shows electron microscopy images from field emission gun scanning of samples at two different image magnifications, which show significant erosion on the surface of air plasma-treated fabric. This alteration in the morphology of the plasma-treated samples (Figure 6(b)) was compared against the control sample (Figure 6(a)).

Scanning electron microscopy-field emission gun images of polyamide fabric at different magnitudes: (a) untreated sample – 10.4 kx and (b) air-plasma-treated sample (surface activation process) – 10.3 kx.
As the literature shows, this erosion is caused by highly energetic and reactive plasma species that increase surface roughness and alter PA 6.6 from hydrophobic to hydrophilic. In addition to the erosion effect, note that there may be breakage in PA 6.6 molecular chains, generating new functional groups or reorganizing existing polymeric groups. 46
Figure 7 shows the analyses of amplified images (from scanning electron microscopy-field emission gun (SEM-FEG)), which show the efficiency of the SiO2 coating process on PA 6.6 fabric.

Scanning electron microscopy-field emission gun images: (a) untreated sample magnified 5000 ×; (b) untreated sample magnified 30,000 ×; (c) treated sample magnified 5000 × and (d) treated sample magnified 30,000 ×.
Particle filtration efficiency, determining differential pressure and fluid repellence, and making masks
Three filters consisting of layers (test specimens, or TSs) were assembled (see Figure 1). Brazil’s ITA (Aeronautics Institute of Technology) ran tests and posted their findings in Test Report No. 1124883-203 dated 10 June 2021.
PFE tests (summarized in Table 2) showed the filtration was efficacious when the average particle diameter was approximately 0.1 μm.
Particle filtration efficiency
Note: particle filtration efficiency tests measure a mask's ability to filter sub-micron particles. Brazil’s ABNT NBR ISO 15052 standard requires the use of 0.105 ± 0.005 µm diameter latex particles.
TS: test specimen.
The results of the breathability aspects of the mask are shown in Table 3. Air replacement pressure (mmH2O/cm2) shows easiness of breathing (breathability). For this test, a certain amount of air passes through the masks and the pressure difference between the two sides is measured. When the difference is small, this indicates that resistance is low, thus making breathing easier. The important point for the high-level mask is the balance between high filtration efficiency and the low-pressure differential.
Pressure differential (ΔP) – breathability
Note: by evaluating the pressure differential, the mask's air permeability is rated per the ABNT NBR 15052:2004 and ABNT NBR 14673:2002 requirements.
Fluid resistance (FR; mmHg) measures resistance to the fluid when it is spread. The pressure is equal to blood pressure (80, 120, 160 mmHg) and passes through the masks, and its penetration on the opposite side of the masks is visually checked. Table 4 shows the results obtained for the masks developed for this study.
Fluid repellence
aPer ABNT NBR 15052, the fluid repellence property is not a mandatory performance requirement and is applicable only to mask models that claim repellent properties.
After treatment, the materials were sent to cutting and sewing companies to make masks per the architecture shown in Figure 1. Several models were tested and adjusted to meet the main requirements per Brazilian Technical Standards. Elastics are of the double-folding type most commonly used for sportswear. Figures 8(a)–(d) show some of the models developed.

The mask models developed: (a) reusable three-dimensional (3D) mask with polyamide fabric finishing and stitching; (b) reusable 3D mask with seamless polyamide fabric finishing; (c) disposable 3D non-woven polypropylene mask with two spunbond–spunbond–meltblown–meltblown–spunbond filters, in which the filters are covered with spunbond–meltblown–spunbond type 60 g/m2 non-woven polypropylene and (d) reusable 3D mask with polyamide fabric finishing and a central fin to prevent mouth contact.
Washing and care instructions
In 2020, the National Health Surveillance Agency (ANVISA) issued a publication with guidelines for using and maintaining face masks, generally made of fabric, which recommends a shelf life of 30 washes. For this, it is necessary to wash the mask to conform to the orientations below:
masks must be washed separately from other clothes and manually; soak in a solution of water with neutral detergent in a proportion of 4 mL per liter of water for 30 minutes; rinse well under running water to remove any detergent residue; avoid twisting the mask tightly and let it dry; ensure the mask is undamaged (in terms of fit, deformation, wear, etc.), or one will need to replace it; store in a closed container.
Conclusions
The washable and reusable face masks produced for this project (Figure 8) were tested by IPT, demonstrating a high PFE, at over 98%, with 2.11 mmH2O/cm2 pressure differential, thus showing good breathability and fluid repellence. The mask filter consists of layers of super-activated carbon fiber fabric and polypropylene non-woven fabrics for dental–medical–hospital use. The filter is sealed in Amni-Virus Bac Off antiviral PA fabric produced by Rhodia-Solvay Group. It is also impermeable to water due to the silica coating synthesized by an innovative plasma process developed for this project. Special features of the masks developed for this project were wearer comfort, reusability and facial self-adjustment allowed by PA antiviral elastic fabric, in comparison with non-woven fabrics used for FFP2 and N95 commercial masks. Hand washing in water with neutral detergent followed by rinsing in running water is recommended. Another special feature is the insertion of the carbon fiber fabric filter that comprises the liquid-adsorption and air-purifier layers; they have a clip and foam for nasal fitting and are sterilized by gamma radiation at the Brazilian Institute of Energy and Nuclear Research’s Radiation Technology Centre (IPEN - CETER) as per certificate NR - CTR-CI-ITA-20210728.
Footnotes
Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
