Screen-printing of chitosan and cationised cellulose nanofibril coatings for integration into functional face masks with potential antiviral activity

Abstract

Masks proved to be necessary protective measure during the COVID-19 pandemic, but they provided a physical barrier rather than inactivating viruses, increasing the risk of cross-infection. In this study, high-molecular weight chitosan and cationised cellulose nanofibrils were screen-printed individually or as a mixture onto the inner surface of the first polypropylene (PP) layer. First, biopolymers were evaluated by various physicochemical methods for their suitability for screen-printing and antiviral activity. Second, the effect of the coatings was evaluated by analysing the morphology, surface chemistry, charge of the modified PP layer, air permeability, water-vapour retention, add-on, contact angle, antiviral activity against the model virus phi6 and cytotoxicity. Finally, the functional PP layers were integrated into face masks, and resulting masks were tested for wettability, air permeability, and viral filtration efficiency (VFE). Air permeability was reduced for modified PP layers (43 % reduction for kat-CNF) and face masks (52 % reduction of kat-CNF layer). The antiviral potential of the modified PP layers against phi6 showed inhibition of 0.08 to 0.97 log (pH 7.5) and cytotoxicity assay showed cell viability above 70 %. VFE of the masks remained the same (~99.9 %), even after applying the biopolymers, confirming that these masks provided high level of protection against viruses.

Graphical abstract

Chitosan and cationized cellulose nanofibril formulations (individually or as a mixture) were screen-printed onto a polypropylene mask layer, and a biopolymer-modified functional layer was incorporated into a three-layer face mask to provide antiviral properties.

1. Introduction

To improve preparedness and combat the next potential outbreak of COVID-19 or other respiratory viral disease successfully, some improvements could be made in the availability and effectiveness of personal protective equipment (PPE), particularly face masks, to defend against microorganisms. Just to prevent the spread of SARS-CoV-2 during the COVID-19 pandemic, the demand for masks increased globally to 4,000,000 per day [1]. Face masks serve the general purpose of providing a type of physical/chemical barrier that protects against airborne pollutants such as pollen, chemical vapours and pathogenic microorganisms [2], [3]. Therefore, the choice of material structure and chemistry is very important. Since SARS-CoV-2 has a size of 60–140 nm, protective textile materials with larger pore sizes have not yet been shown to provide an effective barrier against these viruses or virus-containing droplets [4], [5]. In addition, most currently available face masks are unable to inactivate airborne pathogenic microorganisms, which may be a problem, because SARS-CoV-2 has been shown to be stable on various surfaces for up to 7 days [6]. Therefore, microorganisms adhering to these materials can increase the risk of cross-infection.

After the outbreak of COVID-19, the demand for highly effective antiviral face masks exploded exponentially worldwide [7]. Many strategies have been used to produce novel, better virus protective masks, from pristine masks with improved filtration performance up to mask layers' functionalisation/coating as one of the most attractive approaches. However, to date, most of the antiviral coatings developed on face masks have still been based on metal nanoparticles, such as silver, copper [8], gold and zinc oxide nanoparticles [9], synthetic compounds, such as N-Halamide [3], quaternary ammonium compounds [2], graphene and graphene-based composites [10], and the use of a metal organic framework (MOF) [1], also based on photocatalytic technology [11]. Such coatings also have many disadvantages, among which are their short-term durability, limited efficacy, and, consequently, the development of new resistant bacteria, as the most problematic [12] next to the human and environmental toxicity. Metal-based nanomaterials may enter the human body by inhalation due to their small size and cause harmful side effects [13], the use of photodynamic coatings may affect the antiviral properties of face masks [14], while synthetic polymer coatings may cause toxicity problems due to their degradation in vivo [15], [16].

The growing awareness of the use of human-safe and environmentally-friendly materials as green alternatives to fossil-based polymers has shifted the focus of research to biopolymers [15], [16]. However, to date, few studies have been reported on the use of biopolymer coatings on face masks and filters [17]. Among them, cationic polysaccharides, in particular, chitosan (Ch) and its derivatives, have shown an efficient antiviral activity against human coronaviruses (S protein-polymer complexes) by interacting directly with the virus, resulting in its inactivation [18], [19]. Ch is a natural cationic polysaccharide extracted from shrimp or fungi, and has a number of remarkable properties, such as biocompatibility, biodegradability, non-toxicity, abundant availability and low cost [20], [21], [22]. In addition, modification of Ch enables its applicability in various fields, with improved functional properties [23], [24], [25]. Choi et al. developed a biodegradable, moisture-resistant, highly breathable, and high-performance fibrous mask filter based on cationically charged Ch nanowhiskers [26]. Recently, a written opinion suggested that positively charged electrospun chitosan nanofibres can protect healthcare workers from COVID-19 infections [22]. Among other biopolymers, polysaccharide starch [27], dialdehyde starch [28], gluten [29], cellulose in various forms [30], polyphenol-based [31], [32] and lignin-based [33] extracts, and many others [34], have been also investigated for their potential bacterial and/or antiviral properties when applied to a mask layer. Cationised cellulose nanofibrils (kat-CNF) are another bio-based polymer, which has already proven its antibacterial properties due to the cationised surface [35], [36], formed by a reaction with the quaternary ammonium agents [37]. The antiviral activity of kat-CNF has rarely, if ever, been reported [38], because the focus has been on the antiviral activity of the anionic groups of cellulose in various nano-forms [38], [39]. Moreover, mixture with Ch has not yet been reported as a potential antiviral coating.

In addition to antiviral activity, achieving a mask with good comfort (especially breathability, moisture vapour transport, and waterproofness), as well as high efficiency in removing bioaerosols, are other aspects that must also be considered [40]. In general, the filtering efficiency of the mask is influenced by its specifications and external factors, which include the inherent properties of the materials used in the mask (such as the chemical composition, thickness and packing density of the fibres), as well as the uniformity of the airflow, frequency of the respiration, relative humidity and temperature and loading time [2]. In addition to the coating formulation nature, the surface application technique thus also plays an important role, in order to balance between the antiviral efficacy and comfort properties of the face mask during wearing.

In studies reporting the application of biopolymers to the macro/nanofibres of mask layers, this was done by dip coating/soaking [26], [27], by simple adsorption of solutions or suspensions [31], or by spray coating [33]. However, in these cases, low molecular weight biopolymers were used mostly, so that the breathability was not affected significantly. In our case, where larger macromolecules, as well as cellulose nanofibrils were used as potential antiviral agents, these techniques could lead to the formation of an overall coating, or the coating could not be controlled by thickness and density, so that the pores of the mask layer would be completely blocked. Therefore, the conventional screen printing technique was used to determine the relationship between the airflow and the deposition of biopolymers [41].

For this purpose, all formulations (pure high molecular weight Ch (HMW Ch) and kat-CNF suspensions, and their mixture) were deposited on the inner side of the outer PP layer in the mask, and evaluated further for the coatings' adherence, air permeability, moisture vapour transport, wetting properties, charging behaviour and antiviral activity against the model virus phi6. Additionally, the biocompatibility (in terms of cytotoxicity) was checked of functional PP mask layers towards skin fibroblast and keratinocytes. Subsequently, the biopolymers surface-modified PP layer was integrated into a final three-layered surgical face mask, and tested towards wettability, air permeability, and, ultimately, viral filtration efficiency (VFE). The potential applicability was discussed.

2. Materials and methods

2.1. Materials

The high molecular weight chitosan (HMW Ch; 310–375 kDa, >75 % deacetylated) and glacial acetic acid (AcOH: 99–100 %), were supplied by Sigma-Aldrich, Austria. The NaOH (>98 %) was purchased from Honeywell (Seelze, Germany), while the 2 M HCl was supplied from Carl Roth (Karlsruhe, Germany). The Milli-Q/ultrapure water (resistivity of 18.2 MΩ cm at 25 °C) was prepared using the Milli-Q system (Millipore Corporation, Massachusetts, USA).

The test liquids used in the procedure for evaluating surface free energy (SFE) utilising different calculation models, were Diiodomethane (DI; 99 % purity; Sigma Aldrich Merck KGaA, Darmstadt, Germany), Dimethyl sulfoxide (DMSO; 99.9 % purity; Honeywell Riedel-de Haën, Seelze, Germany), Ethylene glycol (EG; 99 % purity; Fisher Scientific Vienna, Austria), Glycerol (GY; 88 % purity; Alkaloid AD, Skopje, Macedonia) and Milli-Q ultrapure water (W; Millipore Direct 8; Labena Slovenia). The test liquids were used as received.

The chemicals used for the antiviral tests included trypticase soy broth (TSB) (from Sparks, NV, USA), Bacto agar (BD) (from Sparks, USA), MgCl₂ × 6H₂O supplied from Duchefa Biochemie, Haarlem, The Netherlands. The peptone water was prepared from 10 g/L Bacto peptone (BD, USA) and 5 g/L NaCl (Merck, USA).

The chemicals used for cytotoxicity assay were FBS, penicillin, streptomycin, and l-glutamine (Thermo Fisher Scientific, Massachusets, USA), MTT solution (Sigma-Aldrich, MO, USA) and DMSO (Thermo Fisher Scientific, Massachusets, USA).

Water suspended quaternised cellulose nanofibrils (kat-CNF) of 2–3 μm long and highly branched 10–200 nm wide fibrils with around 0.23 degree of substitution, as determined by conductometric titration, and the 40 ± 2 dPa s were prepared by Xylocel Oy, Finland, according to their patent WO/2016/075370 [42] using glycidyltrimethylammonium chloride.

The details of all mask layers (surgical face mask) are given in Table 1 .

Table 1.

The construction parameters of each mask layer.

Layer	Composition	Weight (g/m2) ISO 3801	Thickness (mm) ISO 9237
First/outer nonwoven	Polypropylene (PP)	35 ± 1.8	0.46 ± 0.04
Middle melt-blown filter	Polyethylene terephtalate (PET)	46 ± 2.3	0.22 ± 0.02
Third/inner nonwoven	Polypropylene (PP)	30 ± 1.5	0.21 ± 0.02

2.2. Preparation of biopolymer formulations as coatings and their characterisation

2.2.1. Preparation of chitosan (Ch) and kat-CNF dispersions, and their mixture

An aqueous HMW Ch solution of 2 % (w/v) was prepared by dissolving HMW Ch powder in MilliQ water. Glacial acetic acid (about 5 mL) was added drop-wise during constant magnetic stirring into 295 mL of water with 6 g of HMW Ch, to enable the chitosan's dissolution. The solution was left stirring for 24 h until a homogeneous dispersion was obtained. The pH was adjusted to 4.0. The initial 4 wt% concentrated suspension of kat-CNF was diluted to 0.75 wt%, and homogenised by stirring it with a propeller stirrer (RE 166 Analog, IKA, Germany) for 30 min at 700 rpm. The suspension was adjusted to pH 4 with the dropwise addition of 2 M HCl. The mixture of kat-CNF and HMW Ch was prepared in the same way, where a pre-dissolved 2 wt% HMW Ch solution of pH 4 was added to 1.5 wt% of kat-CNF by stirring with a propeller stirrer for 10 min at 700 rpm, resulting in a final formulation of 1 wt% HMW Ch + 0.75 wt% kat-CNF. All dispersions were evaluated for rotation viscosity at 25 ± 1 °C using a Haake rotational Viscotester V2 (Thermo Scientific, USA) to determine oriented values for evalution of formulations suitability to be used for screen-printing. The 45 ± 3 dPas was determined for 2 wt% acid dispersed HMW Ch, 21 ± 2 dPas for 0.75 wt% kat-CNF and 26 ± 2 dPas for their mixture (i.e. 1 wt% HMW Ch + 0.75 wt% kat-CNF).

To simplify the sample designation, the formulations were abbreviated as HMW Ch (i.e. 2 wt% HMW Ch), kat-CNF (i.e. 0.75 wt% kat-CNF) and HMW Ch + kat-CNF (i.e. 2 wt% HMW Ch + 0.75 wt% kat-CNF).

2.2.2. Scanning electron microscopy (SEM)

The morphology of the kat-CNF was examined by Scanning Electron Microscopy (SEM) using a JSM-IT 800SHL instrument (Jeol, Tokyo, Japan). The kat-CNF dispersion was put onto a double-sided conductive carbon tape with a pipette (about 100 μL), left to dry well, placed on a holder, and sputtered with gold, to ensure conductivity and prevent charging effects. The sample was examined with an accelerating voltage of 3 kV and a variable working distance. The images were acquired using a secondary electron detector.

2.2.3. Attenuated Total Reflectance–Fourier Transform InfraRed (ATR-FTIR) spectroscopy

ATR-FTIR spectroscopy was used to record the elemental composition and specific functional groups of the used biopolymers. All the formulations were dried at 60 °C in a heater, and their infrared spectra were measured using a Pelkin Elmer Spectrum GX NIR FT -Raman spectrometer (Waltham, MA; USA) at an ambient temperature with ATR crystal accessories. The spectra were recorded at ambient conditions in the wavenumber range of 4000–400 cm⁻¹ from accumulating 16 scans at a resolution of 2 cm⁻¹, and with background spectra subtraction. The Spectrum software program was used for the data analysis. All the measurements were carried out at least in duplicate. The acquired spectra were first deconvoluted by automatic baseline corrections, and then by automatic smoothing and ATR corrections. For comparison of individual spectra, all spectra obtained were normalised to 1 at the end.

2.2.4. X-ray photo-electron spectroscopy (XPS)

The chemical composition of all the formulations was determined by XPS using an XPS instrument model PHI-TFA from Physical Electronics (Chanhassen, MN, USA). The solid samples were mounted in the main XPS chamber with an ultimate pressure of 6 × 10⁻⁸ Pa. The XPS was equipped with monochromatic Al Kα1.2 radiation with the photon energy of 1486.6 eV and a hemispherical analyser positioned at 45°. The samples were analysed over an area with a diameter of 400 μm. An electron gun was used for neutralisation of the surface charge during the measurements, and to prevent charging effects. The C-C component in the C1s peak was set to a value of 248.8 eV. The survey scan spectra were measured at a pass energy of 187 eV and with an energy step of 0.4 eV. The high-resolution spectra were measured at a pass energy of 27.35 eV and an energy step of 0.125 eV. The measured spectra were evaluated with the software Multipak (v. 9.6.0.).

2.2.5. Rheological properties

The rheological measurements of the prepared biopolymer formulations were performed at room temperature by using an Anton Paar GmbH MCR 302 Rheometer and Anton Paar RheoCompass software. The (CUO CC27) cylindrical measuring system was used. The rotational flow curve measurements (viscosities vs. shear rate) were performed in a shear rate range from 1 to 100 s⁻¹. The viscoelastic properties were determined by an oscillatory amplitude sweep measurement at a constant angular frequency (10 s⁻¹) by increasing the amplitude from 0.1 to 100 %. The storage (G′) and loss (G″) modulus were determined.

2.2.6. Zeta potential and hydrodynamic size

The Zeta potential (ZP) of all the biopolymer formulations was measured by Electrophoretic Light Scattering (ELS) in a wide pH range with the Anton Paar LiteSizer 500, using an Omega cuvette at a temperature of 25 ± 1 °C. The pH was adjusted automatically with a titration unit, using 0.1 M HCl or NaOH. The hydrodynamic size for all 1 mg/mL formulations at pH 4 was assessed by Dynamic Light Scattering (DLS), using the Omega cuvette and applying the following parameters: the refractive index was 1.5850 and the absorption index was 0.001. The size measurements were performed at 25 ± 1 °C, at a measurement angle of 175° and in automatic quality mode. For each sample, three replicates were performed, and the mean was calculated. The data for ZP and hydrodynamic size were acquired using Anton Paar's Kaliope software.

2.2.7. Surface tension

The surface tension (SFT) of samples was determined using a Wilhelmy plate of known dimensions, according to DIN 53 914 and a K12 tensiometer (KRÜSS GmbH; Germany). By starting the measurement, the drive table with the tested solution (125 mL) rose until it contacted the bottom edge of the Pt plate. The force was measured proportional to the surface tension and the contact angle between the solution and the Pt plate. The K12 Wilhelmy plate setup corresponding software calculated the surface tension automatically by regression of the measurement's points. A more detailed explanation can be found elsewhere [43]. The measurements were repeated 10 times, and are presented as the average value, along with the Standard Deviation.

2.2.8. Antiviral properties

The antiviral properties of all biopolymers formulations towards the model virus, bacteriophage phi6 (often used as a surrogate for SARS-CoV-2 [44]), were determined as already described in [45]. Briefly, HMW Ch, at the final concentration of 1.25 mg/L [45], a mixture of 2.5 mg/mL HMW Ch and 1.9 mg/mL kat-CNF, and 0.5 mg/mL kat-CNF were tested. The first two were tested at pH 4, while the kat-CNF was tested at pH 6.5. The HMW Ch was tested in two replicates, whereas the other two components were tested once.

2.3. Screen-printing procedure and characterisation of mask layers

2.3.1. Screen-printing procedure

The printing (Scheme 1 ) of all formulations was performed on a Zimmer laboratory screen printing machine using a nickel-based rotary screen (SPG Prints B.V., The Netherlands, formerly Stork) of 135 mesh size, 22 % open area, 88 μm of holes' diameter, and 120 μm of screen plate thickness, and a steel-rod type smooth squeegee of 15 mm in diameter, which was pushing the printing formulations through the screen at a relevant pressure (magnet adjustable to no. 2) and rotating the screen at a relevant speed (manual setting, stage 5 ≈ 6 m/min). The mask layers were air dried at room temperature after printing. All the samples were conditioned at standard conditions (21 ± 2 °C and 65 ± 5 % of relative humidity) prior to testing.

Scheme 1.

Schematic presentation of the screen-printing procedure and integration of coatings into a surgical face mask/layer.

2.3.2. Coatings' add-on

The samples` weights were measured and the mass/area (g/m²) was calculated for each sample, and the add-on percentage was assessed by using the following equation (Eq. (1)):

$$\mathrm{Add}-\mathrm{on}\left(\%\right)=\left(m_c-m_i\right)/m_i\times 100$$ (1)

where m _c and m _i are the weights of the sample after and before the treatment, respectively; up to three measurements for each sample were assessed and the average value was reported.

2.3.3. SEM

The SEM images were performed at the same operating conditions as described under 2.2.2. During the SEM imaging, the Energy Dispersive X-ray Spectroscopy (EDXS) mapping was also performed, to determine the presence and distribution of specific elements (i.e. C, N, O).

2.3.4. Surface zeta potential

The streaming potential measurements were performed with a SurPASS 3 (Anton Paar GmbH, Austria) using the adjustable gap cell for sample mounting. A pair of each 20 mm × 10 mm samples was mounted on the sample holder using double-sided adhesive tape. The distance between the samples` surfaces was set to 110 ± 10 μm. The surface Zeta potential was determined as a function of pH in an aqueous electrolyte solution of 10 mM KCl (the ionic strength was high enough to suppress any contribution from pore conductivity) [46], [47], by adjusting the pH automatically with 0.05 M KOH and 0.05 M HCl. Before measurement, the solid sample was equilibrated at a neutral pH with several rinsing steps, and then the pH was adjusted to the alkaline range. A pressure gradient of 200–600 mbar was applied to generate the streaming potential, which was measured with a pair of AgCl electrodes. The pH and conductivity of the electrolyte were monitored with pH and conductivity probes, and all experiments were performed at room temperature. Between each sample analysis, the electrolyte system was rinsed thoroughly with ultrapure water to ensure that all previous solutions were removed.

2.3.5. XPS and ATR-FTIR

The XPS and FTIR analysis of the samples before and after printing were provided as described in Section 2.2.

2.3.6. Contact angle

The sessile drop method and the OCA 35 goniometer (Data Physics, Germany) were used to determine the contact angles on the mask layers. The samples were placed on the solid plate of the apparatus under a stainless-steel needle with an inner diameter of 0.16 mm. A 3 μL drop of Milli Q ultrapure water was released automatically from a glass syringe (500 μL, DataPhysics, Germany) at a dispensing rate of 0.50 μL. A droplet was formed on the sample surface and the contact angle was measured using the Laplace-Young equation. For each sample, 8 replicates were performed (8 contact angles were measured on each sample) on both sides.

The contact angles, besides using MilliQ ultrapure water were also performed with Diiodomethane for the PP layer. First, the arithmetic mean values were calculated, along with the Standard Deviations. Additionally, for a spin coated glass cover coated with the model virus phi6, dimethyl sulfoxide, ethylene glycol and glycerol were also used for surface free energy (SFE) determination. The results were then used by the SCA 21 software program, and by choosing the two mathematical models, i.e., the Owens-Wendt-Rabel-Kelble (OWRK) model [48], [49] and the Wu model [50], [51], the total SFE and its corresponding dispersive and polar contributions were presented in the form of data, and also as a point plot diagram.

2.3.7. Air permeability

The air permeability of the mask layer and laminated mask samples was measured using a Karl Schröder apparatus (Karl Schröder KG, Weinheim, Germany) according to the Standard ISO 9237. The test piece in tension was placed over the clamping head. The value on the scale was read after selecting one among three-grade levels of airflow (5–90, 50–550 and 500–5600 L m⁻² s⁻¹) and turning the air flow regulator to a max value of 1. The sample measurement area was 20 cm². The air permeability of the mask layer and laminated mask samples was calculated using the following equation:

$${\mathrm{V}}_N=f·V_G·\sqrt{\frac{P_U·T_N}{P_N·T_U}}$$ (2)

where V_N is the reference air permeability [L m⁻² s⁻¹], P_N is the reference atmospheric pressure 1013 mbar, T_N is a reference temperature (293 K), V_G is the measured air permeability [L m⁻² s⁻¹], P_U is the measured atmospheric pressure in the environment [mbar], T_U is the measured temperature in the environment [K], f is the factor for corrections depending on the samples' area of measurements (i.e., f = 1 for 20 cm²). The results are presented as average values of 6 measurements along with the Standard Deviation (i.e., the air permeability was measured on 6 different areas within the same sample). Before measurement the mask layers were conditioned at 21 ± 2 °C and 65 ± 5 % of relative humidity.

2.3.8. Water-vapour resistance

The water vapour resistances (WVR) of the samples were assessed according to ISO 11092:2014 using the measuring instrument KES-F7 (Thermo Labo II, Kato Tech Co- Ltd., Japan), where the heat flow was measured through the sample. The wet filter paper was put on the heat plate for evaluating the water vapour resistance (which determines the latent evaporative heat flux across a given area in response to a steadily applied water vapour pressure gradient).

The WVR was calculated from the measured heat flow using the following Eq. (3):

$${\mathrm{R}}_{\mathrm{et}}=\left({\mathrm{p}}_{\mathrm{s}}-{\mathrm{p}}_{\mathrm{a}}\right)·\mathrm{A}/{\mathrm{\phi }}_{\mathrm{et}}$$ (3)

where R_et is the water vapour [Pa m²/W] resistance of the fabric, A is the area of the heat plate [m²], p_s and p_a are the saturated vapour pressures on the heat plate [Pa] and in the tunnel [Pa], and ϕ_et is the corresponding heat flow [W].

The results are given as the mean values of up to ten measurements for each sample.

2.3.9. Antiviral potential

The antiviral activity of the screen-printed mask layers was determined using an enveloped bacteriophage phi6 (DSM 21518). The materials were cut into 5 × 5 cm pieces, and placed in Petri dishes with the screen-printed side up, and bacteriophage suspensions were added in a 1 × SM buffer with the pH value of 7.5 or 4.7. The same was done for the unprinted samples, i.e., the controls. The petri dishes containing the samples were then shaken at room temperature for 2 h. After the incubation period the materials were removed from the Petri dishes, appropriate dilutions of the samples were prepared, and the virus concentrations were determined using a double-layer plaque assay (DAL).

Three media were used for the DAL, e.g. ‘TSB agar’, ‘TSB top agar’ and ‘liquid TSB’. ‘TSB agar’ (30 g/L TSB and 15 g/L agar) was used to prepare the agar plates, ‘TSB top agar’ (30 g/L TSB and 15 g/L agar) was used for the top layer of the double layer plaque assay, and ‘liquid TSB’ (30 g/L TSB) was used for culturing the bacterial culture [Pseudomonas syringae van Hall 1902 (DSM 21482)]. All the media contained 1.93 g/L MgCl₂ × 6H₂O. The DAL was performed by adding 0.2 mL of the bacterial host in a logarithmic phase (prepared by inoculating 0.2 mL of a 2-day-old bacterial culture in 5 mL of ‘liquid TSB’, followed by ≥3 h incubation at 25 °C and 210 rpm), and 0.25 mL of undiluted or diluted virus samples to ~5 mL of melted ‘TSB top agar’ in 15 mL glass tubes. This was then mixed thoroughly, poured onto agar plates and incubated overnight at 25 °C. Afterwards, the number of plaques was counted, and the virus concentration was calculated by considering the virus dilutions and plating volumes as follows:

$$\mathit{Inactivation}\left(\%\right)=\frac{A-B}{A}\ast 100$$ (4)

$$\mathit{Inactivation}\left(log\right)=log\left(A\right)-log\left(B\right)$$ (5)

where A is the virus concentration in the control sample after 2 h incubation, and B is the virus concentration in a sample with antiviral compound after 2 h incubation. Both concentrations were calculated as the average of one or more dilutions, with three technical replicates.

2.3.10. Cytotoxicity of functional mask layers

2.3.10.1. Sample preparation

The samples were extracted at four different concentrations: 100 %, 50 %, 25 %, and 12.5 %, according to ISO10993-12:2012 [52] and ISO10993-5:2009 [53]. Briefly, the materials were extracted at an extraction ratio of 3 cm²/mL in cell culture medium for 24 h at 37 °C and 5 % CO₂. The concentrations of 50 %, 25 %, and 12.5 % were obtained by serial dilution of the 100 % extract with a cell culture medium. The cell culture medium was used as a positive control.

2.3.10.2. Cytotoxicity assay

Skin fibroblasts (Detroit 551 - CCL-110, ATCC) and keratinocytes (a generous gift from Professor Dr. Elsa Fabbretti, University of Primorska), were grown in an Advanced DMEM cell culture medium supplemented with 5 % FBS, 100 IU/mL penicillin, 0.1 mg/mL streptomycin, and 2 mM l-glutamine at 37 °C and 5 % CO₂.

As a quantitative assay, cytotoxicity was assessed using the MTT cell viability assay protocol, which is based on measuring cell viability via metabolic activity [53]. The skin fibroblasts and keratinocytes were seeded in 96-well plates (10,000/100 μL/well) and maintained in a culture. After 24 h the cells were stimulated with the four different concentrations of the extract of the test sample and the cell culture medium. After 24 h of exposure, the cells were incubated with 10 % MTT solution for 3 h at 37 °C. Formazan formation was determined for each treatment concentration, and compared with that of the control cultures. A decrease in the number of live cells resulted in a decrease in the metabolic activity of the cell. Detection of cell viability was performed by quantifying the optical density at 570 nm using a multiplate reading spectrophotometer (Varioskan, Thermo Fisher Scientific, Massachusetts, USA) after removing the MTT solution and then suspending the cells in 100 μL DMSO. Each experiment was performed four times. A viability of more than 70 % of the blank sample was judged to be non-cytotoxic.

2.4. Mask fabrication and characterisation

2.4.1. Lamination procedure and ultrasound welding

The process of laminating the layers into a mask is as follows: All three layers were folded on a folder. The height of the fold was 18 mm. The layers were not glued here, but folded together (they were held together by a fold). This package of layers was then cut to the desired size. The mask was sewn or welded ultrasonically along the edge (inserting the rubber band and aluminium plate into the mask beforehand). Finally, the mask was packed in a plastic package and sterilised in an oven at a minimal 120 °C for half an hour. The depicted procedure of the lamination process is given in Scheme 2 .

Scheme 2.

Schematic presentation of integrating a surface coated outer PP layer into the surgical face mask, the layer lamination and ultrasonic welding, and sterilisation of the mask, including the picture of the final product.

2.4.2. Air permeability

The air permeability was determined in the same way as explained in Section 2.3. Before measurement the face masks were conditioned at 21 ± 2 °C and 65 ± 5 % of relative humidity.

2.4.3. Contact angle

The Contact angle was measured in the same manner as described in Section 2.3.6.

2.4.4. Viral filtration efficiency (VFE)

The Viral filtration efficiency (VFE) tests were performed according to the Standard EN 14683:2010 [54], described in Košir et al. [55], with some modifications. Briefly, peptone water with phi6 was aerosolised, producing droplets and aerosols of different sizes, which passed through the mask (in the case of the positive controls, PCs, there was no material), and were collected in a 6-stage Andersen sampler. Each stage contained a TSB plate, onto which the mixture of 5 mL of melted ‘TSB top agar’ and 200 μL of P. syringae in the logarithmic phase was poured, and this was allowed to harden. After the collections of droplets and aerosols, the plates were incubated overnight at 25 °C, the plaques were counted and the number of viruses was determined considering the positive hole correction [56]. Each experiment consisted of 2 PCs (one was performed at the beginning of the experiments, and the other just before the negative control), 3 control materials, or masks with the same biopolymer formulations (material subsamples), and a negative control (air flowing without viruses, which was performed at the end of each experiment). The VFE was first calculated for each subsample according to Eq. (6):

$$\mathit{VFE}\left(\mathit{SB}\right)\left(\%\right)=\frac{\mathit{Cpc}-\mathit{Csb}}{\mathit{Cpc}}\times 100$$ (6)

where Cpc is the average viral concentration of the two PCs, and Csb is the viral concentration of each material subsample. The final VFE of the PP layers with each biopolymer formulation or a control material was then calculated as the average value of three subsamples according to Eq. (7):

$$\mathit{VFE}\left(\%\right)=\mathit{average}\left[\mathit{VFE}\left(\mathit{SBs}\right)\right]$$ (7)

3. Results and discussion

3.1. Physicochemical and rheological properties of biopolymer formulations

In order to reveal the morphology of the kat-CNF biopolymer, SEM microscopy was applied, where more distinct nanofibrils were observed (Fig. 1 ), and the nanofibrils of the kat-CNF that form the network structure are clearly visible.

Fig. 1.

SEM image of kat-CNF.

Fig. 2a shows the infrared spectra of the HMW Ch, kat-CNF, and their mixtures. The HMW Ch exhibits typical functional groups resulting from its structure. The glycosidic groups at about 1150 cm⁻¹, the band at 1581 cm⁻¹ correspond to N-H vibrations, while the band at 1654 cm⁻¹ belongs to the functional groups -CO-NH. At larger wavenumbers, distinct bands attributed to N-H, O-H, and C-H can be seen, and are present in the HMW Ch structure. Glycosidic functional groups originating from the cellulose structure are observed in the kat-CNF. At 1650 cm⁻¹ there is a band for C-OH groups present on the cellulose backbone. A very distinct band at a larger wavenumber, namely, -OH vibrations are observed, as well as a band for -CH₂. Importantly, quarterly ammonium functional groups are identified at 1481 cm⁻¹, which carry a permanent charge by reacting with the hydroxyl groups with GOPTS reagents [37]. In the infrared spectra of the two mixtures the bands of the functional groups remained in their positions as in the pure individuals. The main difference was the shift of the (CH₃)₃N⁺ functional groups to a lower wavenumber (i.e., 1449 cm⁻¹).

Fig. 2.

ATR-FTIR (a), survey spectra (b) with corresponding high-resolution XPS spectrum of C 1s (c), N 1s (d) and O 1s (e) of biopolymers, and (f) zeta potential as a pH function for all the biopolymers.

Importantly, both biopolymers have cationised groups in their structure (primary amino/HMW Ch and quaternary/kat-CNF), which may play a key role in inactivating model viruses, as highlighted in our recent study [45].

The HMW sample of Ch, kat-CNF and their mixture contained mainly carbon, oxygen and nitrogen, as determined by XPS (Table 2 , Fig. 2b). This composition is typical for Ch, which contains many -OH groups as well as NH₂ groups, which is the main reason for the nitrogen's presence, and especially for the high oxygen content. The rest of the elements may be of impurities' origin (often found in organic samples) with concentrations <1 at. %. The amount of nitrogen in the cellulose itself (kat-CNF) is very low, so the nitrogen in the formulation of both components is originating predominantly from the chitosan, and the additive effect may be seen in both. The most striking differences between the samples were found in the high-resolution nitrogen spectrum of N 1s (Fig. 2d). For the case of kat-CNF, the nitrogen peak is positioned at a higher binding energy of ~402.4 eV, and it belongs to quaternary nitrogen. In the HMW Ch sample, there is no such peak, only the peak characteristic of amino groups (~399.4 eV). However, in the case of the formulation of Ch and kat-CNF, both peaks are observed in the nitrogen spectrum - the quaternary nitrogen is seen as a characteristic knee on the left side of the peak, corresponding to quaternary amino groups. In agreement with the results of ATR-FTIR, the knee for the quaternary amino groups in the mixture has shifted to a lower binding energy. In the carbon C1s spectra (Fig. 2c), C-C, C-O are present for all samples, while O-C-O bonds are present only for a mixture of HMW Ch with kat-CNF, and, to some extent, also for HMW Ch (this is also in accordance with the zeta potential measurements, Fig. 2f), and may result from acetyl groups, whereas no difference is observed in the high resolution O 1s spectrum for all samples (Fig. 2e). Moreover, the proportions of these bonds differ between the samples, in agreement with the differences in oxygen to carbon concentrations, and no other peculiarities are observed.

Table 2.

Atomic surface composition of biopolymers (in at. %).

Biopolymer	C	N	O	Cl	Ca	Si
HMW Ch	70.6	4.6	23.8	/	0.4	0.7
kat-CNF	65.9	1.9	30.4	1.0	0.8	/
HMW Ch + kat-CNF	62.6	5.7	31.0	0.7	/	/

The Zeta potential (ZP) results as a function of pH are shown in Fig. 2f. The HMW Ch, in the form of a macromolecular solution, exhibits a positive charge up to pH 7, which is due to the positively charged amino groups (protonated -NH₃ ⁺). Although chitosans theoretically exhibit IEP at about 6.8, the shift to a higher pH can be attributed to several reasons, including its concentration and chain length [57]. At about pH 7, the amount of protonated -NH₃ ⁺ and -NH₂ is equal, leading to the isoelectric point (IEP). Below the IEP, in a more alkaline range, the HMW Ch reaches a plateau with negative ZP, which is due to the negatively charged residual acetyl groups on the HMW Ch backbone (the degree of deacetylation is >75 %). In contrast, kat-CNF shows no protonation/deprotonation behaviour, which is typical of quaternary ammonium groups. From the ZP curve, a constant positive charge can be observed in the range of 35–40 mV. In the case of the HMW Ch + kat-CNF mixture formulation, the ZP curve is similar to HMW Ch, but differs in the alkaline range. As the alkaline range is approached the positive ZP decreases, while, in the acidic range, a positive ZP occurs at values of about 35–40 mV. The positive ZP started to decrease after pH 4, which follows the deprotonation behaviour of chitosan, and may suggest its dominant behaviour in this mixture, as already shown by the nitrogen content (determined by XPS). Nevertheless, at pH 4 all formulations exhibit a large ZP that indicates their stability clearly.

The results of surface tension, hydrodynamic size and polydisperse index (PDI) of formulations and their antiviral properties are pointed out in Table 3 .

Table 3.

Surface tension, hydrodynamic size, polydisperse index (PDI) and antiviral properties of formulations as potential antiviral agents.

Biopolymer at pH 4	Surface tension (mN/m)	Hydrodynamic size at pH 4 (nm)	Polydisperse index (PDI) in %	Inactivation of phi6 at pH 4.7 (HMW Ch and HMW Ch + kat-CNF) or pH 6.5 (kat-CNF)
2 wt% HMW Ch	58.74 ± 2.81	4703 ± 1860	43 ± 3	99.9999 % or 6.32 log at 1.25 mg/mL
0.75 wt% kat-CNF	87.01 ± 3.69	5480 ± 1970	33 ± 5	84 % or 0.8 log at 0.5 mg/mL
1 wt% HMW Ch + 0.75 wt% kat-CNF	64.24 ± 3.04	8816 ± 2372	32 ± 8	2.5 mg/mL HMW Ch + 1.9 mg/mL kat-CNF 99.999 % or 5.29 log
MilliQ at pH 4	66.70 ± 0.78	/	/	No inactivation

The highest surface tension (SFT) was evident for kat-CNF (Table 3), that could be attributed to larger and stronger nanofibres' interactions in the solution phase, arising from the surface decorated with trimethylammonium quarternary groups (i.e., van der Waals, orientation dipol-dipol interactions, and hydrogen bonds, etc.). If the network structure of differently branched/interlaced fibrils are 10 μm or larger, their weight will cause gravity-induced flotation capillary forces [58] to increase the surface tension. The measured SFT for HMW Ch is lower (i.e., 58 mN/m), as obtained by other authors (i.e., 64 mN/m) [59]. The lower SFT for HMW Ch could be due to the acetic acid added to obtain the pH = 4 [60], and more expressive repulsion between the charged amino groups, which hinders physical interactions. In addition, the SFT for HMW Ch is influenced by its concentration used, since the higher SFT for the increased concentration of HMW Ch could result due a more expressed intermolecular hydrogen bond formation in the solution [61]. It is known that surface tension is higher if bulk interactions are stronger [61]. Only a bit higher SFT was obtained in the mixture of kat-CNF and HMW Ch in comparison to HMW Ch measured SFT (64 mN/m). From these results the dominant character of Ch in the mixture of HMW Ch and kat-CNF is again evident.

In addition, the hydrodynamic size was measured for all formulations at about 0.1 wt% and at pH 4. As shown in Table 3, all biopolymer formulations have a large hydrodynamic size, with the smallest measured for HMW Ch (i.e., 4703 nm at pH 4), followed by kat-CNF, whose size was in the range of about 6 μm. The mixture exhibits a slightly higher hydrodynamic size, possibly to some physical interactions, resulting in an overall larger size of HMW CH + kat-CNF (i.e., 8816 nm). Unusually, the larger polydisperse index (PDI) followed the opposite trend, with the highest PDI observed for HMW Ch, indicating a more heterogeneous size distribution, while the lowest, i.e. 32 %, was measured for HMW Ch + kat-CNF, and may point out the quite nice and equal distribution of fibrils into the chitosan macromolecular solution. Although a rather high ZP value was determined for all solutions at pH = 4 (Fig. 2f) indicating large stability (ZP >20 mV), some agglomeration can obviously occur, leading to a hydrodynamic size in the micrometre range.

Regarding the antiviral properties of the used formulations, it can be summarised from a previous study [45], that HMW Ch shows almost 100 %, or 6.32 log inactivation potential, towards phi6 at a concentration of 1.25 mg/mL, while kat-CNF shows only 84 %, or 0.8 log inhibition, at 0.5 mg/mL. Similarly, a high inactivation potential of 99.9999 %, or 5.29 log, was determined for the mixture of both components, probably due to the inhibitory effect of HMW Ch (Table 3).

The rheological properties of the prepared biopolymer formulations were measured in shear rates and stresses to simulate their printing behaviour, i.e. shear-thinning with high enough viscosity at higher shear stresses and pseudo-plastic behaviour after its reduction. For a successful printing process it is important that the printing suspension can be transferred to the substrate easily, and is able to recover structurally after being squeezed through the screen openings by forming uniform and homogeneously distributed patterns on the material's surface. After the paste passes through the screen meshes elasticity becomes desirable, in order to prevent slumping or any major changes in shape [62]. In this context, the viscosity in shear was measured for all biopolymer formulations. Fig. 3a shows the shear viscosity as a function of shear rate, indicating the highest viscosity at low shear rate for the kat-CNF suspension (23682 ± 991 mPas), and the lowest viscosity for the HMW Ch solution (1306 ± 63 mPas). All samples exhibited a shear thinning behaviour, from which the highest decrease in viscosity over increasing shear rate was observed for the kat-CNF sample, which was expected behaviour for suspensions containing flexible physically entangled fibrils' structures [63], which tend to orient and partly disentangle by the flow, leading to a lower viscosity as the shear rate increases [64]. Comparing the individual HMW Ch solution and kat-CNF dispersion with their mixture (HMW Ch + kat-CNF) a slight change in rheology behaviour is observed, with a small increase in viscosity at low shear rate (2033 ± 203 mPas) and a desirable increase in negative slope, indicating a better shear thinning behaviour in comparison with the pure HMW Ch solution.

Fig. 3.

Rheological properties of the biopolymer formulations.

Fig. 3b shows the loss and storage moduli as a function of the shear strain for all three samples. It was observed that the storage modulus was higher than the loss modulus (G′ > G″) only for kat-CNF, indicating its predominant elastic behaviour and a crossover or flow point (G″ > G′) where the sample began to flow, indicating that the fibrillar network structures had broken down and began to behave as a non-Newtonian shear-thinning fluid. A larger difference between storage and loss modulus was observed for the HMW Ch solution, which showed a predominant liquid-like behaviour with almost linear and strain independent response in the lower shear strain regions and their sudden drop [65]. Overall, it seems that the addition of the kat-CNF dispersion in the prepared formulation mixture (HMW-chitosan with kat-CNF) improved the rheological properties favourable for the screen-printing process.

3.2. The effect of screen-printed biopolymer formulations on the properties of the inner side of the PP non-woven layer

The SEM images with smaller magnification (180×) to obtain a larger view of the applied screen-printed layer showed that the HMW Ch filled the pores of the layer (marked with arrows) to some extent after being applied to the PP mask layer (see Fig. S1 in Supplementary Material). It should be noted that the white spots are the result of the calendering process, which determines the mechanical properties of the mask layer. When applying kat-CNF to the PP layer, the presence of the biopolymer becomes visible as a coating that covers the PP microfibres partially, and some are covered completely by the layer. When a blend of both biopolymers was applied, the density of the pristine mask layer covered by the formulations was greater. It can be seen that, in some cases, the patterned coatings cover the voids in the nonwoven mask layer to some extent, which, in turn, affects the air permeability, as shown later.

In addition, the morphology of the pristine and biopolymer formulations screen-printed PP mask layers was examined with higher magnification, in combination with EDXS elemental mapping (Fig. 4 ). The pristine PP mask layer shows smooth fibres with a size of about 20 μm, which are superimposed randomly. Consistent with the composition of PP, EDXS mapping showed only the uniform distribution of element C (no presence of N and O). After the introduction of the structured layer of HMW Ch, some coatings can be seen on the PP fibres, as well as in the pores of the layer. By performing the mapping analysis, typical elements such as C, N and O were confirmed, and indicated the presence of the applied Ch formulations layer. The clear distribution of N dots (blue dots) indicated the uniformity and patterning of the applied HMW Ch pattern. In the case of the printed kat-CNF, the spot-deposited biopolymer covers individual fibres by being imprinted even in the empty voids between the fibres. At higher magnification (see Fig. S2 in the Supplementary Material), the presence of a nanofibrillated cellulose mesh covering the PP microfibres is clearly visible. Similarly, the EDXS mapping confirmed the components typical of kat-CNF and their distribution among the PP fibres. Based on the N-mapping, it was again possible to infer the distribution of the screen-printed biopolymer layer. A similar conclusion can be drawn for the PP layer with screen-printed HMW Ch + kat-CNF, where the mapping confirmed all the elements present and their distribution, indicating the presence of both introduced biopolymers.

Fig. 4.

SEM images and EDXS mapping (elements C, N, O) of the pristine PP non-woven first mask layer, and the corresponding layer with screen-printed HMW Ch, kat-CNF and HMW Ch + kat-CNF.

Deposition of the screen-printed formulations onto the PP nonwoven layer was also confirmed using ATR-FTIR, and their imprinting/penetration by surface zeta potential analysis measured on both surface sides of the nonwoven PP (see Fig. 5 ).

Fig. 5.

ATR-FTIR spectra of the pristine and screen-printed PP non-woven with biopolymers (a) and the surface zeta potential of the PP layers' surfaces before and after the screen-printing of HMW Ch (b), kat-CNF (c) and HMW Ch + kat CNF (d).

The deposition of the biopolymers' formulation on the PP nonwoven was also confirmed by ATR-FTIR spectroscopy (Fig. 5a). Typical bands, such as -CH₃ at 1463 cm⁻¹, -CH₂ at 1369 cm⁻¹, and -CH at about 2900 cm⁻¹, were observed, consistent with the structure of PP. When the HMW Ch was applied, a new band appeared at 1050 cm⁻¹, corresponding to the C-O-C of HMW Ch. In addition, new bands at around 3000 cm⁻¹ wavenumbers indicated the presence of O-H and N-H functional groups. Similar spectra were observed after the deposition of kat-CNF, where more pronounced peaks at about 1370–1470 cm⁻¹ and 2800–3000 cm⁻¹ were also observed due to the C-H bending and stretching of the methyl groups. The sample printed with the mixture of HMW Ch and kat-CNF showed a similar trend. The band at around 2355 cm⁻¹ corresponds to CO₂ from the air [66].

The surface zeta potential (SZP) measurements, performed as a function of pH, are supporting the ATR-FTIR analysis. The reference PP sample, analysed on both sides (inner onto which the formulations were screen-printed, and outer), show the typical behaviour of PP. Above the isoelectric point (IEP, i.e., pH of the aqueous solution at which ZP = 0 mV) at about pH 4–5, the PP-water interface was negatively charged, whereas the sign of this interfacial charge reversed to positive at low pH. This behaviour is typical for a polymer or other hydrophobic material surface without surface functional groups [67], but specific adsorption of ions from the liquid media. Above pH 4–5, the adsorption of hydroxide ions dominated, and produced a negative interfacial charge, as evidenced by the negative zeta potential [67]. At pH 4–5, the concentration of adsorbed OH⁻ ions was equal to that of H₃O⁺ ions accounted into the IEP. Below pH 4–5, the specific adsorption of H₃O⁺ ions dominated, making the interfacial charge positive. The difference between the sample's inside and outside could be due to structural / constructional changes, leading also to different wettability (see the contact angle results in Table 4 ; small differences can be seen), which is also reflected in the differences in the plateau value of the negative SZP (from pH 8). In the case of the HMW Ch screen-printed sample (see Fig. 5b), the IEP shifted to a lower pH value, from about 7 for pure HMW Ch (Fig. 2f) to about 5.7 for PP coated with HMW Ch, due to the presence of the PP layer. Moreover, after HMW Ch coating, the positive charge of the PP layer acquired higher values, and the positive SZP plateau value increased from 10 mV (PP) to 25–30 mV, additionally confirming the attachment of HMW Ch to the PP. It can also be observed that the PP surface became less hydrophobic after the deposition of polar HMW Ch, as the negative ZP plateau values decreased. Comparing the HMW Ch-printed side with the outer/non-printed, it can be seen that the coating penetrated through the microporous structure of the PP nonwoven, given the almost the same positive SZP plateau values; the IEP was shifted only by one unit further to the lower pH. This was to be expected, as Ch is in a fully soluble form at pH 4, leading to a stretched conformation, and, thus, easier passage through the pores of the material. A similar behaviour was observed for the screen printed kat-CNF dispersion (Fig. 5c). After the deposition of kat-CNF, the IEP of the PP shifted to a higher pH (6.4), which was related to the quaternary amino groups as weak bases. This was also reflected in a higher positive SZP below the IEP compared to the reference samples. The rather high reduction of the negative SZP values was due to the hydrophilic character achieved by the deposition of kat-CNF, which also penetrated to the other side of the PP layer through its microporous structure, but to a much lesser extent than in the case of HMW Ch. This was to be expected, because HMW Ch in a macromolecular solution is smaller in size than kat-CNF that has 2–3 μm long and highly branched 100–300 nm wide and interconnected fibrils (Fig. 1). The contribution of HMW Ch was more pronounced when the mixture of HMW Ch and kat-CNF was screen printed on the PP nonwoven (Fig. 5d). This can be illustrated by the very similar curves of ZP = f (pH) for both samples, i.e. HMW Ch + kat CNF and HMW Ch (Fig. 2f).

Table 4.

Summarised data of the screen-printed non-woven PP mask layer.

Screen-printed layer	Add-on (%/m2)	Water vapour retention (Pam2/W)	Air permeability at 100 Pa (L/m2s)	Contact angle – printed side (°)	Contact angle – non-printed side (°)	N atomic composition at the PP nonwoven layer (at.%)	Surface zeta potential at pH 7.5 on the printed side (mV)	Surface zeta potential at pH 7.5 on the non-printed side (mV)	Surface zeta potential at pH 4.7 on the printed side (mV)
PP_HMW Ch	3.40 ± 0.05	5.814 ± 0.047	8.5 ± 1.0	130 ± 8	125 ± 9	1.0	−30	−30	20
PP_HMW Ch + kat-CNF	2.34 ± 0.02	5.738 ± 0.065	25.4 ± 2.0	131 ± 8	140 ± 9	3.0	−22	−41	23
PP_kat-CNF	0.91 ± 0.01	5.602 ± 0.062	45.0 ± 4.1	107 ± 14	133 ± 8	1.5	−5	−32	12
Pristine PP 1 mask layer	/	5.845 ± 0.141	103.5 ± 3.7	139 ± 12	134 ± 15	0	−53	−38	−5

The effect of biopolymer formulations` deposition was also evident from the contact angle values, add-on and air permeability, as shown in Table 4.

The largest add-on (Eq. (1)) was determined for PP printed with HMW Ch, which accounted for an increase of 3.4 %/m² compared to the original mask layer (reference). A slightly lower add-on was observed for the biopolymers mixture (i.e., 2.34 %/m²), and the lowest value for the kat-CNF suspension (0.91 %/m²), also corresponding to the initial concentrations of the formulations.

The air permeability results followed the same trend. The reference material showed the highest value of air permeability, which was decreased after deposition of the different formulations, The smallest decrease (43 %) in air permeability was observed for the kat-CNF printed sample, followed by using the mixed formulation (i.e., HMW Ch + kat CNF), while it was reduced even up to 92 % by printing of HMW Ch alone, which indicated their different structuring on the nonwoven. The kat-CNF showed the highest viscosity, but also good shear-thinning and pseudo-plastic behaviour (see Fig. 3), which means their structural recovery after elastic deformation, resulting in more surface deposited and thin-layer patterned structures on the nonwoven, and where its imprinting was affected primarily by the micro-porosity of the nonwoven and pressure of the squeegee; these statements can be also supported by the highest surface tension (87.01 mN/m) of the kat-CNF suspension, which, in addition to its physical stability and good water retention properties [63], are critical for the mask's applicability. On the contrary, the rheological non-elastic Newtonian HMW Ch solution with much lower surface tension (58.74 mN/m) spread after deposition on the entire surface and diffused into the fabric, filling the pores, thus, giving the highest add-on and lowest air-permeability (Fig. 4, Fig. S1 – SEM results). The mixture formulation with a surface tension of 64.24 mN/m and moderate shear-thinning and viscoelastic behaviour reflects the properties and behaviour of both biopolymers.

The CA values, measured on both sides of the PP nonwoven layer, additionally supported these results. The highest CA was observed for the reference sample PP, known to be hydrophobic due the presence of non-polar CH and CH₂ groups along its backbone, and a CH₃ pendant group. Taking into account a screen – printed side, the CA was reduced (by 23 %) only after the deposition of the kat-CNF formulation, which, however, still indicated hydrophobic behaviour, while keeping the other (unprinted) side unchanged – similar to the reference. The application of HMW Ch and its mixture with kat-CNF showed almost the same CA values to the reference sample, while the CA of the mixture on the opposite non-printing side increased by 6 %.

The water vapour resistance (WVR) values were relatively low and showed no obvious difference between the reference layer PP (i.e., 5.814 ± 0.047 Pam²/W) and the biopolymer-modified layer (Table 4). The durability of the applied coatings was evaluated by mimicking human breath, and the weight variation of the functional samples was 0.01–0.05 %, confirming the durability and stability of the screen-printed coatings (see Supplementary Material for details).

An important conclusion here can be that, with the selected coatings, we did not change the hydrophobicity of the outer surface of the PP nonwoven dramatically, which is very important for practical use. Hydrophobicity may be effective in repelling statically dispensed aerosols droplets of viruses. Moreover, breathing in the mask produces moisture condensation, which, in the case of a hydrophilic mask, would cause the mask to become wet and shorten its usability or usefulness [68]. A quite good correlation between the WVR and CA was shown, with a correlation coefficient of R² = 0.8624 (see Fig. S3 in Supplementary Material). To understand the interaction between the biopolymers and the PP better, the SFE of the PP layer was calculated and related to the SFT of the biopolymer solutions (see Table 3, first column). The SFE was calculated according to the Owens, Wendt, Rabel, and Kaelble (OWRK) and Wu models, as described in detail in [69]. According to the OWRK method, the total SFE for PP is 3.17 mN/m, and according to the Wu method, it is 5 mN/m. The first model shows the dispersive part in the range of 3.17 mN/m and no polar part, while the Wu model gives 5 mN/m of the dispersive part and 0.5 mN/m of the polar part. These results support the well-known fact that the PP material has a low surface free energy (3.17 mN/m according to the Wu method), and the dispersive part dominates. PP represents typical surfaces with low surface energy due to its hydrocarbons structure. This material has very low reactivity with poor wetting affinity, and due to the relatively high SFT of all biopolymer formulations, the adsorption affinity of these biopolymers to PP is rather low, with hydrophobic and Van der Waals (London) interactions dominating – which may be the highest for the kat-CNF bearing hydrophobic methylated functional groups.

Fig. 6 shows the antiviral activation of the differently screen-printed surface PP layer at the printed side, tested at the lower pH of 4.7 and higher pH of 7.5, which mimics the pH of exhaled breath condensate (pH 7.5) [70] and salivary droplets (pH 6–7) [71]. This was performed in order to understand more clearly the mechanism of action and behaviour of active groups present on the mask surface. In terms of add-on, the screen-printed layer showed no particular correlation with antiviral activity at two different pH values (Fig. 6). Despite the fact that HMW Ch and HMW Ch + kat-CNF deposited on PP also had a relatively high CA (hydrophobicity) that somehow prevented the model virus from wetting the coated mask layer, the inactivation at pH 4.7 was greater than that of PP_kat-CNF, where there was no inactivation, even though this surface had the lowest CA (107°, Table 4). It has been shown previously that viral persistence on material can also be influenced by surface hydrophobicity [68], [72]. In particular, the hydrophobicity of the outer protein layers in viral capsids can influence interactions with solid surfaces and the environment [72]. Thus, a basic understanding of the interactions between viruses, i.e., their outer surface (in the case of phi6 or other enveloped viruses it is made of lipids and proteins), and solid surfaces, is critical for controlling environmental transmission and developing efficient antiviral strategies. Both experimental and computational analyses were performed, to determine the hydrophobicity of various viruses, and it was concluded that the sorption of hydrophobic viruses, i.e., viruses with hydrophobic protein outer layers, is favoured by surfaces coated with hydrophobic sorbents, whereas sorption of hydrophilic viruses is favoured by hydrophilic surfaces [73].

Fig. 6.

Add-on and antiviral potential of the biopolymers screen-printed surface of the PP mask layer.

Bacteriophage phi6 is composed of three layers, and the outer layer is made of phospholipids [74]. The latter exhibit both hydrophilic and hydrophobic properties. The SFT, its polar and dispersive components and polarity were determined, to understand the interactions of PP and PP-coated masks better, even in a water (buffer)-based phi6 dispersion. For phi6, the total SFT in the buffer solution was 51.9 mN/m, with a polar component of 15 mN/m and a dispersive component of 36.9 mN/m. Thus, the inactivation at pH 4.7 was greater for HMW Ch and the mixture of HMW Ch + kat-CNF than for kat-CNF, where no inactivation occurred. This was due to the fact that this mask layer (kat-CNF) had a lower contact angle of 107° (Table 4), and to the dominant dispersive fraction of phi6, which may hinder the wettability of suspension with phi6, therefore preventing the contact of viruses with the porous structures and the antiviral coating of the mask. It should be noted that surface roughness has also led to different surface wetting behaviour, which, in turn, affects the bioactivity of a surface [72]. The presence of porous structures and a coating surface morphology affects the roughness of a surface coating, which also alters antiviral activity [75]. In general, the applied formulations in the form of patterns still had a large hydrophobicity (Table 4), which limited the contact of the model virus with the antiviral component due to the lower wettability.

Finally, microbial adsorption on surfaces is regulated mainly by the most important forces, namely, electrostatic interaction and Van der Waals interaction, where the charge state of a surface and the microbes affects the adsorption rate of the microbes [75]. Therefore, the antiviral activity was strongly dependent on pH, as pH influences the charge of biopolymers (coatings) strongly, as already shown in our recent article [45]. At a slightly acidic pH (i.e., 4.7), the inactivation of phi6 followed the SZP as a surface charge indicator on the screen-printed side. This was confirmed with the greatest inactivation activity for the mixture deposited on the mask layer, which was 1.63 log at an SZP of 23 mV (Table 4). Here, the atomic N composition was also the highest (3 atomic %), and originated from different amino groups (primary and quaternary). On the screen-printed side, a slightly lower SZP was measured at pH 4.7 for PP_HMW Ch (20 mV), and the inactivation for phi6 was about 1.1 log. Surprisingly, despite the positive SZP of 12 mV at pH 4.7, no inactivation activity was observed for the printed side with kat-CNF applied. Obviously not enough amino groups were available on the surface for successful inhibition. Quaternary amino groups may also be steric hindrances for viral phospholipids to come too close and interact electrostatically. A completely different behaviour for the inactivation of phi6 on mask layers was observed at pH 7.5. While PP_HMW Ch hardly showed any inactivation (0.08 log), only a slightly higher value was determined for the applied mixture (0.47 log). These low values are the result of deprotonated amino groups and lower accessibility of quaternary ones. The dominant character of HMW Ch in the mixture as a coating has already been shown from the physicochemical parameters, and the same, of course, applies to the bioactivity. However, comparing the later data with the SZP, it is clear that both modified mask layers showed a negative SZP at a comparable pH (Table 4). On the other hand, the mask layer with the applied kat-CNF alone showed an inactivation activity of 0.97 log towards the model virus at pH 7.5, but also with an SZP of being slightly negative (−5 mV, Table 4). The slightly negative SZP can be due to the fact that the kat-CNF was deposited in the form of patterns (SEM images, Fig. 4 and Fig. S1), and the overall SZP can originate not only from kat-CNF, but also from the PP not being fully coated (only partially), indicating the presence of a negative charge from the PP. The antiviral activity of kat-CNF can be attributed to both cationic groups and hydrophobic moieties, which may additionally contribute to the antiviral potential [37], as well as to a denser nonwoven structure (mechanical effect) that entraps viruses physically compared to pure PP microfibres [7]. In addition, the inhomogeneity of the applied layers may also have a negative effect on antiviral activity to some extent (Figs. 4 and S1 in Supplementary Material).

It can be concluded that several important surface properties, such as surface wettability, surface morphology and surface chemical properties, are important, and should be taken into account when making an antiviral coating, to make it more durable and long-lasting. Among them, electrostatic interactions are obviously one of the most important. It has been shown that protonated amino groups contribute better to antiviral efficacy. This could be due to the ionic interactions between the negatively charged shells of phi6 and the protonated amino groups. This is also supported by the fact that the phospholipid bilayer of bacteriophage phi6 consists of the protein P9, whose amino acid composition contains up to 11 mol% glutamic acid [76]. However, it is difficult to focus on only one parameter, because the combination of parameters may be responsible for the antiviral activity.

The cytotoxicity of the functional screen-printed samples was also determined quantitatively using an MTT assay on the cell viability of skin fibroblasts and keratinocytes; the results are shown in Fig. 7 . Cells plated with the appropriate growth medium (control medium) were healthy. After 24 h of extract admission, all the samples passed with viability >70 %, as recommended in ISO 10993-5:2009 [53]. The ISO Standard describes materials as toxic as soon as their extract leads to >30 % cell death. Moreover, the ISO 10993 Standard was created to ensure the safety and biocompatibility of medical devices (facial masks). For surface products on the skin, the Standard recommends evaluating their cytotoxicity. The Standard states that a product can be considered non-cytotoxic if the highest concentration of its extract (100 %) allows cell viability above 70 % after 24 h of exposure [53]. Our studies showed that the functional PP layers with screen-printed HMW Ch, kat-CNF and their mixtures as coatings did not exhibit cytotoxicity, indicating that these materials are suitable for the development of facial masks.

Fig. 7.

Effect of functional PP layers with screen-printed HMW Ch and kat-CNF coatings on the cell viability of skin fibroblasts (a) and the effect of functional PP layers with screen-printed HMW Ch and kat-CNF coatings on the cell viability of keratinocytes (b).

3.3. The effect of a biopolymer coated inner side of PP non-woven on the comfort properties of laminated surgical face masks

In Table 5 , the results of the CA, air permeability and VFE of laminated surgical face masks (as shown in Scheme 2) are presented, depending on the used formulation added on the first inner-surface layer by screen-printing.

Table 5.

Air permeability, Contact angle and VFE of the modified laminated masks.

Modified laminated mask	Air permeability VN (L/m2s)	Contact angle (°)		VFE
		Outer mask side	Inner mask side
Pristine mask	69	156.33 ± 3.62	130.42 ± 5.47	99.976 %
Mask with HMW Ch	13	151.83 ± 7.06	137.71 ± 6.97	99.967 %
Mask with kat-CNF	33	147.35 ± 4.38	145.58 ± 3.30	99.89 %
Mask with HMW Ch + kat-CNF	32	147.29 ± 7.26	137.34 ± 4.33	99.94 %

The reference sample showed the highest air permeability, being comparable to the report by Čepič et al. [77]. The air permeability of a mask lowered by 33 % when using a pristine PP nonwoven (Table 4) integrated into the mask. Among the used formulations, the HMW Ch coated PP layer gave the lowest air flow through, and the reduction was >80 % compared to the reference (Table 5). The other two formulations used (i.e., kat-CNF and HMW Ch + kat CNF) showed the same air permeability results, while the reductions compared to the reference laminated mask were 52 % and 54 %, respectively. Yet, considering the airflow according to the Standard for face masks EN14683 [54] that passes through the mask, it should be >8 L/min [78]. Taking into account the measured air permeability, the value for the unmodified mask accounted for 8.3 L/min, which fulfils the requirement, while for the most promising case, kat-CNF (i.e., 33 L/m²s, Table 5), a value of 4 L/min was obtained, which is twice less than recommended. With the exception of the reference mask, none of the face masks modified with biopolymer exceeded the mentioned required value of airflow. It should be pointed out that the most representative parameter to characterise the mask according to the EN14683 Standard [54] is the pressure drop, which is a measure of breathability. For Type I masks, the differential pressure should be <40 Pa/cm² at the required measurement conditions.

The Contact angle, as a measure of wettability, was measured on both sides of the face masks, and is shown in Table 5. On the outer side of the mask, the Contact angle decreased after application of the biopolymers compared to the reference value. This was due to the fact that the biopolymers also penetrated through the layer, as already demonstrated by the SZP (Fig. 5). Surprisingly, the Contact angle increased slightly on the inside of the mask (facing the front). Nevertheless, both laminated face masks had high hydrophobicity, indicating good rejection of water viruses` droplets (saliva, breath condensate, etc.) with integrated viruses, and allowed consumer-friendly use, because the mask did not get wet during use while breathing.

Incorporation of the screen-printed layer with any biopolymer into the final three-layer mask did not affect the VFE, which was above 99 % in all combinations (Table 5). This means that these masks provided the same filtration efficiency and, thus, the same protection, as the standard three-layered masks without biopolymers. A VFE measurement is, however, not a Standard method for evaluation of the filtration efficiency of masks, but, as we described in our previous work, VFE is comparable to bacterial filtration efficiency (BFE), which is the Standard method [54]. The data on VFE show that masks with polymer addition fulfil the requirements of the Standard EN14683:2019 to some extent, and can be used safely for personal protection against virus infection via aerosols.

While VFE meets the recommendations (Table 5), airflow through the laminated mask has been reduced significantly because the pores were filled with coating substrates, and thus the required level of airflow was not achieved. Although the individual layers showed some level of antiviral potential, and, thus, positive potential (Fig. 6), it is still far from the recommendations (approximately a reduction of 2–3 log). This can be attributed to the lack of positive charge at the pH of salivary droplets, as well as exhaled breath condensate. Therefore, a possible option would be to use lower molecular weight (MW) and shorter antiviral bio-based molecules with a permanent positive charge - the charge density should be greater without affecting the respiratory capacity significantly. Nanotechnology could also be of help here. For example, Ch particles of equal and controlled size could avoid pore filling and allow air permeability. The mode of deposition also plays a key role. While we have used a highly controlled deposition of biopolymers, another potential application could be electro/spraying with a controlled droplet size and uniform distribution of the latter. On the other hand, the PP SFE energy should be increased by material activation, such as plasma treatment. This could improve the affinity of the coating to the PP. Thus, proposed possible solutions in the future could be:

• I)Use of smaller MW bio-molecules with a permanent positive charge or nanoparticles, and enhance their deposition (add-on) to increase antiviral activity, but still remain within the range of required breathability
• II)Using a different deposition method that does not take the form of a full coating (spraying, electro-spraying)
• III)Use of plasma treatment to hydrophobic PP microfibres to provide new functional groups for better adhesion of coatings
• IV)Achieving excellent antiviral activity and acceptable breathability already with the first layer as a modified layer, as this determines the further VFE and airflow in the final laminated mask.

4. Conclusion

In the present work, the biopolymers HMW Ch, kat-CNF and a mixture thereof were screen-printed successfully onto the first PP layer in the form of a thin patterned layer. The morphology after the application of all the biopolymer formulations to the first PP mask layer showed their presence, which was also confirmed by infrared spectroscopy. The surface zeta potential (SZP) results provided additional evidence for all the applied formulations, and also showed a partial transition of the formulations to the outer surface of the PP layer. The add-on to the mask layer was the greatest for HMW Ch. The air permeability decreased after the application of the biopolymers, and the smallest reduction was observed for the kat-CNF formulation. As for the antiviral activity of the modified PP layer, it was strongly pH-dependent and the highest for the mixture deposited on the mask layer, with 1.63 log at an SZP of 23 mV at an acidic pH. Here, the atomic N composition was also the highest (3 atomic %), and came from the different primary and quaternary amino groups. On the other hand, when mimicking the real conditions, the inactivation was worse, as the PP_HMW Ch screen-printed layer hardly showed any inactivation (0.08 log), and only a slightly higher value was obtained for the applied mixture (0.47 log), while kat-CNF alone showed an inactivation activity of 0.97 log. In the case of the Contact angle (CA, hydrophilic/hydrophobic properties), only the kat-CNF formulation reduced the CA compared to the reference sample, and it was concluded that wettability altered the antiviral assay and its activity to some extent. The cytotoxicity was also determined quantitatively using an MTT assay for the cell viability of skin fibroblasts and keratinocytes, and showed no cytotoxicity of the functional screen-printed samples.

The formulations applied to the first layer additionally worsened the ability of airflow throughout the laminated mask, but still kept the mask hydrophobic on both sides, which is very important to eliminate aerosol droplets with integrated viruses and offers good wearability to the user (wet masks are unfriendly to the user). The data on VFE showed that masks with polymer addition fulfilled the requirements of the Standard EN14683:2019 to some extent, and thus have a potential to be used for personal protection against virus infection via aerosols.

CRediT authorship contribution statement

Olivija Plohl: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Visualization, Writing – original draft, Writing – review & editing. Vanja Kokol: Conceptualization, Investigation, Methodology, Supervision, Visualization, Writing – original draft, Writing – review & editing. Arijana Filipić: Formal analysis, Investigation, Methodology, Writing – review & editing. Katja Fric: Formal analysis, Investigation, Methodology, Writing – review & editing. Polona Kogovšek: Funding acquisition, Formal analysis, Investigation, Methodology, Writing – review & editing. Zdenka Peršin Fratnik: Formal analysis, Investigation, Methodology, Writing – review & editing. Alenka Vesel: Formal analysis, Methodology, Writing – review & editing. Manja Kurečič: Formal analysis, Investigation, Methodology, Writing – review & editing. Jure Robič: Formal analysis, Methodology, Writing – review & editing. Lidija Gradišnik: Formal analysis, Investigation, Methodology, Writing – review & editing. Uroš Maver: Funding acquisition, Writing – review & editing. Lidija Fras Zemljič: Conceptualization, Funding acquisition, Resources, Project administration, Supervision, Writing – original draft, Writing – review & editing.

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.

Appendix A. Supplementary data

Supplementary material

Data availability

Data will be made available on request.