Polylactic acid as a suitable material for 3D printing of protective masks in times of COVID-19 pandemic

PLA material and masks preparation by 3D printing

Polylactic acid (PLA) was purchased in the form of filament for FDM 3D printing from Shenzen Creality 3D Technology Co., LTD, China. Protective masks, circular plates (diameter of 10 cm and height of 0.2 cm, printed vertically) and square carriers (1 × 1 cm, 0.2 cm high) (Fig. 1) were prepared using a 3D printer (Prusa i3 MK3, Czech Republic). The printing template was designed with Trimble Sketchup Pro, exported in a stereolithographic (stl) file (freely available at https://www.facebook.com/groups/1346383268879783/files/) and used to print the objects of investigation. The printing parameters were as follows: layer height = 0.3 mm, shell thickness (perimeter) = 0.4 mm, bottom/top thickness = 0.2 mm, fill density = 10%, print speed = 90 mm/s, extrusion temperature = 215 °C, platform temperature = 60 °C, filament flow = 95%, machine nozzle size 0.4 mm, the infill pattern was grid (i.e., linear tilted 45°) and the total layers = 338.

Visualization of PLA mask structure using scanning electron microscopy

The structure and porosity of PLA 3D-printed masks were examined using a scanning electron microscope (SEM) Nova NanoSEM 450 (Fei, USA). Approximately 1 × 1 cm pieces cut from printed masks were completely air-dried and visualized by SEM. Since the material is very sensitive to electron exposure, mild conditions had to be used, that is, voltage of 5 kV and low vacuum. Images of each visualized position were captured by LVD detector at gradual magnifications 2,000×, 1,000×, 500×, 100× (focusing on identical position), dwell time 5 µs and spot size 4.5. The size of the pores between PLA filaments was measured and marked using a SEM operating software (xT microscope Control v6.3.4 build 3233), provided by the SEM manufacturer (Fei, USA). The SEM images shown in this study were selected as representative visualizations of the PLA microstructure. The SEM analysis merely illustrates the 3D-printed PLA surface morphology, and gap width measurements are not analyzed statistically.

Visualization of PLA mask structure under stress conditions using scanning electron microscopy

To investigate the impact of possible stress factors for PLA masks, cleaning with chemicals was performed and exposure to wearing-associated contamination was simulated, as outlined below.

The effect of immersing in three chemical disinfectants (96% ethanol, 70% isopropanol and the commercial disinfectant and bleach SAVO Original, Unilever ČR s.r.o., Czech Republic containing 0.85% sodium hypochlorite diluted with water (2:9)) was tested by repeated (5 × 15 min) cycles and long-term (24 h) exposure.

The simulation of human impact on the PLA structure was performed as follows: extensive exposure to fingers (to simulate incorrect application of the mask), abrasion with paper (minor mechanical stress) and dining fork (strong mechanical stress), immersion in 1.9% sodium chloride solution for 4 h (to simulate perspiration).

Short rinsing with 100% acetone was examined to investigate its effect on PLA surface properties. After each treatment, completely dried PLA carriers were examined using SEM, as described above.

Aerosol particle passage through PLA material

Surface contamination (and potential surface penetration) with infectious agents was simulated by an aerosol generator and a photometer (Lorenz Meβgerätebau FMP 03) with a differential pressure sensor (Fig. S1), providing a suitable test system with stand for facial masks and flat filter materials. The device was certified as a test system according to the standards EN 143 (Respiratory protective devices—Particle filters—Requirements, testing, marking), and EN 149 (Respiratory protective devices—Filtering half masks to protect against particles—Requirements, testing, marking). A sample of the PLA material was attached in the standardized testing cartridge and was sealed with silicone to prevent false positive detection of penetrating particles passing along the edge of the PLA panel (Fig. S2). The cartridge was mounted into the Lorenz Meβgerätebau FMP 03 device, between the aerosol generator and photometer. The aerosol generator produced a defined amount of aerosolized paraffin oil, the test system passed it through the material, and the photometer situated on the other side of the PLA sample measured the aerosol concentration, thereby indicating the retention efficiency. An integrated differential pressure sensor was used to determine the pressure loss during passage through the sample. The particle size distribution was approximately 0.1–2 µm (geometric mean 0.44 µm), which is close to the most frequently observed penetrating particle size (Fig. S3). The output of the aerosol generator was set to 150% with flow 95 L/min, atomizer pressure 5 bar and oil temperature 60 °C. The test was performed for 270 s.

Disinfection of PLA material artificially contaminated with bacteria and yeast fungus

Wild strains of S. epidermidis, E. coli and C. albicans were used as representatives of gram-positive and gram-negative bacteria or yeast fungus, respectively. The concentration of bacteria was adjusted to approximately 1 × 10⁷ colony forming units (CFU) per mL, the fungus concentration was 1 × 10⁶ CFU/mL. Each PLA carrier with a size of 1 × 1 cm was contaminated with 10 µL of microbial suspension applied to the surface of carriers in 1 µL droplets for 1 h. The disinfection of contaminated carriers was carried out by immersing in three mL of 96% ethanol, 70% isopropanol, or 0.85% sodium hypochlorite (SAVO Original, Unilever ČR s.r.o., Czech Republic) for 15 min. After evaporation of disinfectant solutions, the carriers were immersed in one mL of sterile 0.9% saline, vortexed, and the obtained suspensions were inoculated onto appropriate agar plates. Blood agar was used for S. epidermidis, Müller-Hinton (Oxoid, Czech Republic) agar for E. coli and Sabouraud agar (Oxoid, Czech Republic) for C. albicans. Samples not exposed to treatment by disinfectants were used as controls. The inoculated plates were incubated at 37 °C for 48 h. Each experiment was done in triplicate, and results were obtained by counting the average CFU/mL.

Disinfection of PLA material artificially contaminated with viruses

SARS-CoV-2, the causative agent of the COVID-19 pandemic, was isolated in a biosafety level 3 laboratory from a nasopharyngeal swab by inoculating Vero CCL81 cells (ECACC 84113001) and subsequent expansion by two additional passages in Vero CCL81 cells. Passage 3 was cleared by centrifugation at 1000 g for 5 min, passed through a 0.45 µm filter, and stored at −80 °C until use. In addition to SARS-CoV-2, inactivation of a stable DNA virus, the Human Adenovirus 2 ATCC VR-846 (HAdV) obtained from the American Type Culture Collection (ATCC) was assessed.

Similar to the previous set of experiments, PLA carriers of 1 × 1 cm size were contaminated with 20 µL of a SARS-CoV-2 suspension displaying a median tissue culture infectious dose (TCID50) of 10⁶ IU/mL, which was applied to the surface of carriers in 1 µL droplets. An additional set of carriers was covered with 50 µL of HAdV suspension (10⁶ virus copies) spread evenly over the entire surface. The contaminated carriers were then immersed in 96% ethanol, 70% isopropanol or 0.85% sodium hypochlorite for 15 min. Subsequently, residual viruses—if present—were washed from the dried surface using 180–200 µL PBS. The solution was used directly for infection of Vero-E6 cells (ATCC CRL-1586), in case of SARS-CoV-2, or A-549 human lung carcinoma cells (DSMZ ACC107 from German Collection of Microorganisms and Cell Cultures), in case of HAdV, respectively. Recovered SARS-CoV-2 was titrated by an immunofluorescence (IF) assay using a 1:2.5 serial dilution of Vero-E6 cells starting from 10 µL. Vero-E6 cells were incubated for 72 h at 37 °C in a CO₂ incubator prior to the IF assay. Briefly, medium was washed out, cells were fixed using 4% paraformaldehyde (PFA), cell membranes were perforated with 0.2% Triton-X100, and SARS-CoV-2 was labeled with primary mouse anti-SARS-CoV-2 antibody. Secondary anti-mouse antibody was conjugated with a Cy3 fluorophore and a fluorescent microscope (Olympus IX 81, Germany) was used for signal detection. In the case of HAdV, serial dilutions of virus inoculum were used to infect A-549 cells and the cytopathic effect (CPE) was determined using Motic AE21 Inverted Phase Contrast Microscope (Zeiss, Germany). The titers of both recovered viruses infection particles were determined as TCID50 and calculated using the Spearman-Kärber method (Kärber, 1931; Spearman, 1908). In addition, recovered HAdV genome copies were determined by real-time quantitative PCR (RQ PCR) as described previously (Lion et al., 2003) using the ABI Prism Fast 7500 Instrument (Thermo Fisher Scientific, MA, USA).

Disinfection of PLA masks worn by test persons

To investigate the feasibility of disinfecting PLA protective masks in practical use, three volunteers wore the protective masks of the same type for 4 h. Thereafter, smears from one half of the inner (approximately 80 cm²) or outer surface (approximately 83 cm²) of each mask were performed using sterile cotton swabs. These samples served as a control for natural mask contamination by manual handling, direct skin contact and exhalation. Each cotton swab was transferred into one mL of 0.9% saline in a microtube, vortexed and inoculated onto a blood agar plate. Thereafter, the filters were removed from masks and the PLA skeletons of the masks were immersed in 96% ethanol for 15 min. After ethanol evaporation, cotton swab smears were taken from the second halves of the inner and outer mask surfaces, inoculated onto agar plates, and incubated at 37 °C for 48 h. The results were averaged and expressed as CFU/mL.

Investigation of structure and porosity of 3D-printed PLA material

The structure and porosity of PLA masks produced by 3D printing were investigated by SEM. Scanning electron micrographs of gaps between the PLA filaments were captured at four different magnifications (Fig. 2). The PLA filament size determined was 312.8 µm (Fig. S4) and its surface appeared macroscopically very smooth (Fig. 2A). Further magnification showed only slight roughness of the surface and very small gaps between filaments (Figs. 2B and 2C). Additional increase of magnification revealed connecting filaments of PLA, resulting from the high temperature during 3D printing, with only very small pores (6.049 µm in size) in between. The pores appeared to be completely closed deeper in the carrier, as observed at the highest magnification used (2,000×) (Fig. 2D). To further test whether the pores were indeed closed and prevented particles from passing through the printed mask, we determined the number of paraffin oil aerosol particles displaying a size of 0.1–2 µm using the aerosol generator and photometer, certified as a test system according to the common standards. Maximum pressure loss of the generated aerosol was detected, and absolutely no penetration occurred even though the PLA sample was printed with a diameter of 10 cm (corresponds approximately to the printed height of the masks) in the vertical position, simulated printing at a lower temperature in the upper layers (on the z-axis).

Effect of ethanol, isopropanol and sodium hypochlorite on disinfection of PLA material contaminated with bacteria, yeast fungus or viruses

The results of disinfection of artificially contaminated PLA are summarized in Tables 1 and 2. Although the untreated PLA carriers were contaminated by highly concentrated bacterial suspensions of 1 × 10⁵ CFU/mL, complete decontamination by all disinfectants used was achieved. Single colonies were observed in the samples of S. epidermidis and E. coli disinfected by isopropanol, but these isolated findings can reasonably be considered a contamination that occurred after treatment of the samples. The disinfection of PLA carriers contaminated with C. albicans (4 × 10⁴ CFU/mL) was complete in all cases.

Titers of SARS-CoV-2 and HAdV recovered from disinfected or untreated carriers were determined by IF- and CPE-based assays, respectively. All disinfection agents tested showed complete virucidal effects against SARS-CoV-2. Disinfectants per se exhibited a cytotoxic effects on Vero-E6 cells (Table S1), but this effect was eliminated by serial dilutions during virus titer determination. HAdV infectivity was reduced by ethanol and isopropanol, and completely abolished by sodium hypochlorite. Similar trends were observed by RQ-PCR performed for detecting the HAdV genome copy numbers (Table S2).

Investigation of PLA structure after exposure to ethanol, isopropanol and sodium hypochlorite

The effect of disinfectants on the PLA structure was investigated using SEM. PLA structure, gaps between filaments, and the structure of pores after five 15 min cycles of immersing the carrier in different disinfectants are shown in Fig. 3. Treatment with ethanol (Fig. 3B) resulted in slight melting of the PLA filaments, as compared with untreated PLA (Fig. 3A). The overall PLA structure and surface did not change, but, interestingly, the gap size between the filaments was reduced from the original 6 µm to approximately 850 nm (Fig. 3B). This indicates that ethanol treatment may improve the PLA mask properties with regard to structure density. Similarly, isopropanol treatment did not significantly affect the PLA structure (Fig. 3C). Only slight melting was detectable, resulting in decreased gap size to 3.3–4 µm, in comparison to 6 µm in control samples. Moreover, the surface of filaments remained undamaged. Figure 3D depicts the effect of sodium hypochlorite, which did not alter the surface of filaments, but precipitated disinfectant filled the gaps between them, while the gap size remained almost the same as in the control sample (5–7 µm).

Long-term treatment of PLA by immersion in disinfectants for 24 hours was also investigated using SEM (Fig. 4). The effect of long-term treatment with ethanol (Fig. 4B) was similar to repeated exposure to sodium hypochlorite (Fig. 3D), that is, the gaps between filaments were significantly enlarged to 23.84 µm (Fig. 4B), possibly filled with etched polymer. Investigation of aerosol particle passage through the PLA material after 24 hours in ethanol confirmed that the enlarged gaps were sealed, as no penetration was detected. PLA melting was also observed after prolonged isopropanol treatment (Fig. 4C). The gaps between filaments were sealed with the polymer in an irregular manner, resulting in variable gap sizes ranging from 1.3 to 4.1 µm. As in all previous tests with ethanol, the surface of PLA filaments remained unaffected. In contrast, long-term treatment with sodium hypochlorite damaged the surface of PLA filaments and revealed precipitation of the disinfectant on the surface (Fig. 4D). Similarly to short treatment with sodium hypochlorite, the gaps between filaments, ranging from 2 to 3.5 µm, were completely filled with precipitated sodium hypochlorite (Fig. 3D).

Disinfection of PLA masks by ethanol upon wearing by test persons

To complement the results of disinfection upon artificial contamination (Tables 1 and 2), disinfection of PLA masks after natural use was investigated. The disinfection efficiency with ethanol (96%) is summarized in Table 3. The microbial load detected on the inner surface of untreated masks varied significantly between different users, ranging from hundreds to thousands CFU/mL. Despite this variation, an average of 7 CFU/mL remained detectable after immersing the masks in ethanol for 15 minutes (short rinsing with ethanol was not sufficiently effective; Fig. S5). On the outer surface of untreated masks, 50–150 CFU/mL were detected, and an average of 2 CFU/mL remained detectable after disinfection (Fig. 5).

Visualization of PLA structure upon mechanical and chemical challenge

The impact on the PLA material by finger contact, abrasion by paper or metal and treatment by sodium chloride solution (mimicking perspiration) was analyzed using SEM (Fig. 6). Although fingers may be greasy or sweaty, the contact did not cause any marks or alterations on the PLA surface (Fig. 6A). Similarly, gentle mechanical abrasion with paper did not affect the material (Fig. 6B). By contrast, intensive mechanical scraping with a dining fork significantly damaged the PLA structure (Fig. 6C), leading to compression of PLA filaments, reduction of inter-filament gaps, and shedding of PLA pieces (Fig. 6D). However, neither loosening of filaments, nor increase in gap size or other deformations were observed. Soaking in sodium chloride solution did not affect the structure, but salt crystals were present in the gaps between filaments (Fig. 6E). In addition, the effect of acetone, which is known to damage PLA, was evaluated. Virtually no gap was visible between filaments upon treatment, indicating that even short exposure to acetone smoothens the structure and seals the pores (Fig. 6F).