Evaluation of Antimicrobial Effect of Zinc Pyrithione against Airborne Fungi and Bacteria Growth Collected onto New and Loaded HVAC Fibrous Filters

: Microbial growth onto HVAC ﬁlters was observed in real conditions with possible degradation of the indoor air quality. The ﬁltration performance of marketed antimicrobial ﬁlters containing zinc pyrithione was tested under laboratory conditions and compared to that of similar ﬁlters with the same classiﬁcation, F7 (EN779:2002). The ﬁltration performance of the two tested ﬁlters during loading with PM10 particles was quantiﬁed in an experimental setup with ﬁlter pressure drop measurement and particle counting upstream and downstream of the ﬁlters. The microbial growth on the new and loaded ﬁlters, both contaminated with a microbial airborne consortium composed of two bacteria (Gram-positive and -negative) and fungi, was quantiﬁed by colony-forming units after conditioning the ﬁlters for a few days under controlled temperature (25 ◦ C) and humidity (50% or 90% relative humidity). The results reveal that there was no degradation of the ﬁltration performance of the ﬁlters treated with the antimicrobial agent. The efﬁciency of the antimicrobial treatment, i.e., the ability to inhibit the growth of microorganisms during the incubation period, was signiﬁcant with the new ﬁlters regarding the fungal growth, but the results demonstrate that the antimicrobial treatment became inefﬁcient with the loaded ﬁlters. with used ﬁlters compared to unused ﬁlters. The same ﬁlters as in the present study were tested over 7 months in realistic conditions with semi-urban outdoor air [26]. The results conﬁrmed the antimicrobial effectiveness of zinc pyrithione regarding fungi cultivated with DRBC agar, i.e., that the fungal concentration on the regular PP ﬁlter was signiﬁcantly higher than that on the ZPT/PP ﬁlter, with no inﬂuence of the loading of the ﬁlter considering a moderate increase in the ﬁlter pressure drop by about 30 Pa. No conclusion was established regarding the bacteria because of the high statistical deviation.


Introduction
Indoor air pollution refers to a variety of chemical, biological and physical agents that pose a health threat or cause discomfort. Airborne particles or aerosols are responsible for about 5-34% of all indoor pollution, and their abundance depends on indoor moisture and temperature conditions [1,2]. Bioaerosols are essentially composed of allergens, bacteria, molds, fungi, spores, endotoxins, mycotoxins and all types of particles produced by living organisms with highly variable and complex characteristics. Bioaerosols and, more specifically, microbial aerosols are of major concern in office and commercial buildings because of the close relationship with health diseases (episodes of asthma, rhinitis, conjunctivitis, etc.) [3].
The number of people exposed to bioaerosols in indoor environments can be reduced by appropriate air renewal or air cleaning processes such as filtration, ultraviolet germicidal irradiation, photocatalytic oxidation [4] or plasmacluster ions [5]. For example, the effectiveness of negative ions was investigated on the survival of Serratia marcescens and Staphylococcus epidermidis in an air duct flow [6]. Despite the existence of advanced treatment technologies for bioaerosol inactivation, the most common technique implemented in air handling units of heating ventilation and air-conditioning (HVAC) systems remains filtration with fibrous filters, allowing the collection of bioaerosols, although leading to a possible microbial contamination of the filters with microbial growth.
Microbial growth onto HVAC filters was observed in the literature under laboratory conditions [7][8][9][10], and also in real conditions [11,12], with possible degradation of the Processes 2021, 9,1528 2 of 13 microbial indoor air quality. To address this issue, innovative air filters with photocatalytic properties [13] or containing antimicrobial treatments are designed to be implemented in HVAC systems, or in portable room air cleaners or automobiles. The antimicrobial agents are various, e.g., nanoparticles of copper or silver, or carbon nanotubes [14]. Many natural antimicrobial products have been discovered; e.g., Sophora flavescens nanoparticles revealed an inactivation efficiency against S. epidermidis between 35 and 72% according to the level of humidity [15]. The activity of Melaleuca alternifolia oil (tea tree oil) against a range of microorganisms including Staphyloccocus aureus, Escherichia coli, Candida albicans and Aspergillus niger was shown [16]. The antimicrobial activities of 44 pure natural compounds and two derivatives were determined with only 23 compounds effective in inhibiting the growth of the tested organisms (S. aureus, Bacillus subtilis, Escherichia coli, Pseudomonas aeruginosa and Candida albicans) [17].
The antimicrobial efficiency of the coating or treatment varies according to the antimicrobial agent and, in particular, may depend on whether the filters are new or loaded. The main issue with antimicrobial air filters is that the coating may lead to changes in the filtration performance [18,19], particularly by increasing the airflow resistivity of the filter.
Zinc pyrithione (ZPT) with the chemical formula C 10 H 8 N 2 O 2 S 2 Zn, also known as Zinc Omadine or Zinc 2-pyridinethiol-1-oxide, is used to prevent microbial degradation and deterioration of manufacturing starting materials such as plastics, polymers and latexes, and in a wide range of finished articles made from these starting materials. This chemical acts to prevent the growth of bacteria, fungi, mildew and algae that can cause various types of deterioration such as discoloration, staining and odors. ZPT was found to have a safety profile and/or antimicrobial efficacy that exceeded iodine, chlorhexidine gluconate and triclosan [20]. It acts as an ionophore, interacting nonspecifically with the plasma membrane and shuttling copper into the cell. There is a precedent for pyrithionemediated ionophore activity across intracellular membranes, as it was previously reported that pyrithione affected zinc transport across vacuole vesicles in vitro [21]. The results of most studies on fungi showed that ZPT inhibits fungal growth through damaging ironsulfur clusters of proteins essential for fungal metabolism and also by depolarizing the cell membranes and preventing membrane transport.
The aim of this study was to investigate, under laboratory conditions, the antimicrobial efficiency of a marketed filter containing zinc pyrithione which is known for its effect against microorganisms. The filtration performance of the filter, i.e., the particle collection efficiency and the pressure drop, and the antimicrobial effectiveness regarding a bacteriafungi consortium, i.e., the ability to inhibit the growth of microorganisms during the incubation period, were evaluated and compared to the performance of a regular filter with the same classification. The effect of the level of humidity during filter conditioning as well as the presence of filter loading with organic particles was investigated.

Fibrous Filters Tested
Two multi-layer polypropylene (PP) fibrous filters marketed by Lydall, both of them containing electrostatic charges and classified as F7 according to the European standard EN 779:2002, were studied: a PP filter containing antimicrobial treatment with zinc pyrithione (ZPT) named the PP/ZPT filter, and a regular PP filter without antimicrobial treatment. Few antimicrobial HVAC filters are currently commercially available, perhaps because of the potential toxicity or negative effect of some antimicrobial agents, such as triclosan, which is suspected of increasing antibiotic resistance. The antimicrobial filter with ZPT selected for this study is one of the few commercially available filters.
F7 class filters, in particular, the 2 tested filters, are widely used in the air handling unit of HVAC systems of commercial or office buildings in the first or second treatment stage. The multi-layer PP filters provide high filtration efficiency and dust holding while minimizing airflow resistance, but the fibers lose their charge due to particle loading and air humidity, reducing the efficiency of the filters. The two filters were each composed of five layers: a spunbond layer (SP1), a meltblown layer (MB1) containing the zinc pyrithione in the case of the PP/ZPT filter, a meltblown layer (MB2), a meltblown layer (MB3) containing pink pigments (reference color for F7 class efficiency), a spunbond layer (SP2). The two SP layers ensure the mechanical resistance of the filters, and there is an increasing gradient of efficiency between the MB1 and MB3 layers. The MB1 layer of the PP/ZPT filter is obtained by fusion of regular PP granules and granules containing ZPT, extrusion and then spinning.
The structural properties of the two filters were quantified (PP/ZPT filter vs. PP filter): the thickness (Z) was 1.41 ± 0.08 mm vs. 1.46 ± 0.12 mm (averages and standard deviations for N = 15 quantified by micrometers); the porosity was 91.1 ± 0.3% vs. 89.0 ± 0.8% (averages and ranges for N = 2 quantified by mercury intrusion with mercury porosimeter Autopore IV 9500); the fiber median diameter of the most effective filtering layer (i.e., MB3) was 5.1; 2.9 µm vs. 4.5; 2.6 µm (median diameters; standard deviations for N = 100 fibers analyzed by picture analysis with ImageJ software from scanning electron microscopic observations with EOL JSM-5800LV); the basic weight was 136 g/m 2 vs. 139 g/m 2 (obtained from the manufacturer). The airflow resistance (K) of the new PP/ZPT and PP filters was quantified from filter pressure drop (∆P, Pa) measurements vs. filtration velocity (ϑ) in a ventilated duct with airflow control: where µ is the air dynamic viscosity considering the conditions of temperature and humidity during the tests. The two types of filters were tested in a flat geometry in this study. The airflow resistance was 3.9 ± 0.1 × 10 7 m −1 vs. 3.5 ± 0.1 × 10 7 m −1 (averages and ranges for N = 2) for the PP/ZPT filter vs. PP filter. The results indicate a higher airflow resistance of the PP/ZPT filter compared to the PP filter. In particular, the PP/ZPT filter displayed a higher pressure drop than the PP filter by around 10% at the nominal filtration velocity of 0.12 m/s (recommended by the manufacturer), i.e., 84 Pa vs. 77 Pa. The airflow permeability of the filters (B, m 2 ) was calculated from K considering Darcy's law for laminar airflow: The airflow permeability of the filters was 3.6 ± 0.2 × 10 −11 m 2 and 4.1 ± 0.4 × 10 −11 m 2 for the PP/ZPT filter and the PP filter, respectively. Note that the range in the airflow permeability data is higher, due to the high experimental error in the thickness measurement.
The 2 tested filters display equivalent porous structural properties, but they are not similar.
The quantity of zinc in the filters was quantified at the laboratory by an atomic absorption spectrometer (AAnalyst 200, PerkinElmer) after calcination of filter samples at 550 • C for 4 h and acid attack with HCl (3N). The quantity of zinc measured was 0.15 ± 0.02 ppm for the PP/ZPT filter and 0.19 ± 0.02 ppm for the MB1 layer of the PP/ZPT filter, while there was no quantifiable value for the PP filters (i.e., below the quantification limit of 0.09 ppm). Thus, considering the basic weight of the PP/ZPT filter, the mass of zinc per surface of filtration was around 20 µg/m 2 .

Water Retention Capacity
The microbial growth onto fibrous filters is influenced by the air humidity [8], but even more by the water retention capacity of the filters [8,10]. Water is sorbed by the fibers of the filter and/or by the particles collected by the filter according to the hydrophilic characteristics of the fibers and the particles.
The water retention capacity of the tested filters was quantified by a Karl Fisher device (870 KF Titrino plus, 803 Ti Stand, 860 KF Thermoprop, Metrohm) following the methodology described by [10], after conditioning in an airtight container (see Section 2.3.4) for two levels of humidity exposure: 25 • C and 90% RH (i.e., 21 g of water per m 3 of air), and 25 • C and 50% RH (i.e., 11 g of water per m 3 of air). The water retention capacity was Processes 2021, 9, 1528 4 of 13 expressed in mg of water per surface of filtration. The average values were compared by using the statistical t-test.

Study of the Microbial Behavior on the Tested Filters
The antimicrobial effect of ZPT was tested with new filters, i.e., filters with just microbial contamination representative of the antimicrobial effect in the first operating days, and with loaded filters, i.e., loaded filters with organic and microbial particles. The tests were performed with the PP/ZPT filter and the PP filter with 3 filter samples to ensure the reproducibility of the results.

Preparation of the Microbial Consortium
The microbial contamination of the filters was ensured with a bacteria-fungi consortium, consisting of Staphylococcus epidermidis cells (CIP 53 124), Serratia marcescens cells (DSM30121) and Penicillium chrysogenum spores (wild strain).
S. epidermidis is a bacterial species found on the skin, and thus it is frequently present in indoor air from office buildings in particular. S. marcescens is a motile, short rod-shaped, Gram-negative, facultative anaerobe bacterium, classified as an opportunistic pathogen, and thus its presence in air could cause some pneumonia and urinary tract infections. It is capable of producing a pigment called prodigiosin, which ranges in color from dark red to pale pink, depending on the age of the colonies [22]. P. chrysogenum is a species from the type Penicillium often found in indoor air [23]. Many protease allergens from different species of Penicillium, including P. chrysogenum, are implicated in the occurrence of symptoms found in occupants of contaminated buildings [24].
The culture medium used for the microbial growth was a liquid nutrient broth (Biokar diagnostics) composed of tryptone (10 g/L), meat extract (5 g/L) and sodium chloride (5 g/L) for S. epidermidis and S. marcescens, while soy-based Rose Bengal Chloramphenicol agar (Biokar diagnostics) was used for P. chrysogenum spore production. After their growth, S. epidermidis and S. marcescens cells were collected in an isotonic solution (NaCl 9 g/L), while P. chrysogenum spores were collected in peptone water on the surface of the agar plate. All the harvested strains were washed three times with isotonic solution and by centrifugation for 10 min at 2000× g.
The microbial consortium was prepared on the same day as filters were contaminated. The 2 bacterial species were harvested at the beginning of the stationary growth phase.

Filter Contamination with the Airborne Microbial Consortium
The microbial contamination of the tested filters with the airborne bacteria and fungi was performed in a vertical filtration column ( Figure 1). The microbial suspension was aerosolized with a pneumatic nebulizer, AGK 2000 (Palas). The experimental setup was composed of a straight stainless-steel cylindrical pipe of 45 mm in diameter and 1.5 m in length. The tested filter was located in the middle of the column in a removable filter holder. The airflow through the tested filter was ensured by the nebulizer connected at the top. The microbial contamination was performed at the nominal filtration velocity of the filters, i.e., 0.12 m/s, with the filters tested in a flat geometry. The microbial suspension composed of the 3 species simultaneously was aerosolized upstream of the filters for 10 min at a flow rate of 6 L/min.

Filter Loading with Organic Particles
The loaded filters were obtained under laboratory conditions by clogging new filters with organic particles of micronized rice with a size distribution representative of PM10 particles. The micronized rice was selected to provide nutrients for microbial growth. The clogging was carried out in the same vertical filtration column as previously described for microbial contamination using a rotating brush particle generator (RBG 1000, Palas) operating at a rate of 5 m 3 /h for 2 min. The average generated particle concentration was around 1.5 g/h. The micronized rice particles naturally contain the fungus P. chrysogenum, with a concentration of 2250 ± 850 CFU/g of micronized rice (colony-forming units) [9]. The particle size distribution of the micronized rice particles was characterized with the optical particle counter WELAS 2100 (Palas) with a mass and a number median diameter of 7.4 µm and 0.6 µm.
After particle loading, the filters were contaminated with the microbial consortium in the filtration device as explained previously for the new filters. Note that the quantity of P. chrysogenum coming from the micronized rice during the 2 min of loading was estimated to a maximal value of 10 2 CFU, i.e., assuming negligible loss of cultivability with generation and a negligible deposit in the setup. This quantity was negligible compared to the CFU coming from the microbial suspension nebulized upstream of the filter (10 12 CFU).

Filter Incubation and Quantification of Microbial Concentration Extracted from the Filters
Once contaminated by the microbial consortium, the new or loaded filters, still located in their filter holder, were incubated in an airtight container with controlled temperature and humidity for 8-10 days at 25 °C and 50% RH or 90% RH ( Figure 2). The two humidity conditions were selected because 50% RH is the setpoint in HVAC systems, and a previous study [8] showed that filter conditioning at 90% RH ensured significant microbial growth onto the filters. During the conditioning period, the filters were no longer subjected to airflow. Ventilation stops in HVAC systems are quite common during nights, weekends or vacations. The time period of 8-10 days corresponds to a deteriorated situation representative of a ventilation stop during a 1-week vacation period, and it was selected to ensure significant microbial growth quantifiable with the culture-based method.

Filter Loading with Organic Particles
The loaded filters were obtained under laboratory conditions by clogging new filters with organic particles of micronized rice with a size distribution representative of PM10 particles. The micronized rice was selected to provide nutrients for microbial growth. The clogging was carried out in the same vertical filtration column as previously described for microbial contamination using a rotating brush particle generator (RBG 1000, Palas) operating at a rate of 5 m 3 /h for 2 min. The average generated particle concentration was around 1.5 g/h. The micronized rice particles naturally contain the fungus P. chrysogenum, with a concentration of 2250 ± 850 CFU/g of micronized rice (colony-forming units) [9]. The particle size distribution of the micronized rice particles was characterized with the optical particle counter WELAS 2100 (Palas) with a mass and a number median diameter of 7.4 µm and 0.6 µm.
After particle loading, the filters were contaminated with the microbial consortium in the filtration device as explained previously for the new filters. Note that the quantity of P. chrysogenum coming from the micronized rice during the 2 min of loading was estimated to a maximal value of 10 2 CFU, i.e., assuming negligible loss of cultivability with generation and a negligible deposit in the setup. This quantity was negligible compared to the CFU coming from the microbial suspension nebulized upstream of the filter (10 12 CFU).

Filter Incubation and Quantification of Microbial Concentration Extracted from the Filters
Once contaminated by the microbial consortium, the new or loaded filters, still located in their filter holder, were incubated in an airtight container with controlled temperature and humidity for 8-10 days at 25 • C and 50% RH or 90% RH ( Figure 2). The two humidity conditions were selected because 50% RH is the setpoint in HVAC systems, and a previous study [8] showed that filter conditioning at 90% RH ensured significant microbial growth onto the filters. During the conditioning period, the filters were no longer subjected to airflow. Ventilation stops in HVAC systems are quite common during nights, weekends or vacations. The time period of 8-10 days corresponds to a deteriorated situation representative of a ventilation stop during a 1-week vacation period, and it was selected to ensure significant microbial growth quantifiable with the culture-based method.
After conditioning, the microbial concentration in the filters was quantified by colonyforming unit (CFU) counting. Microorganisms were extracted from the conditioned filters according to the following methodology [8]: filter stirring for 1 h in 50 mL of extraction solution (MgSO 4 (0.01 mol/L), Tween-20 (0.25%)), and then ultrasound for 1 min. The application of this protocol allowed recovery between 80 and 85% of the microorganisms present on the filter media.
Extracted solutions were then diluted and spread on 3 different plates according to the microbial species: -P. chrysogenum was cultivated on Rose Bengal Chloramphenicol agar [25], and the plates were incubated at 25 • C for 5 days; -S. epidermidis was cultivated on Chapman agar culture medium, and the plates were incubated at 37 • C for 24 h; -Serratia marcescens was cultivated on nutrient agar culture medium, and the plates were incubated at 30 • C for 48 h.
Finally, bacteria and fungi colonies were counted, and concentrations were expressed in colony-forming units (CFU) per filter surface. The limit of quantification (LQ) was set to 10 CFU/cm 2 .
The possible presence of Zn in the extracted solution, adversely affecting the plating (CFU) results, was evaluated by comparing the microbial concentrations extracted from new and loaded PP and PP/ZPT filters right after contamination, i.e., at t = 0. After conditioning, the microbial concentration in the filters was quantified by colony-forming unit (CFU) counting. Microorganisms were extracted from the conditioned filters according to the following methodology [8]: filter stirring for 1 h in 50 mL of extraction solution (MgSO4 (0.01 mol/L), Tween-20 (0.25%)), and then ultrasound for 1 min. The application of this protocol allowed recovery between 80 and 85% of the microorganisms present on the filter media.
Extracted solutions were then diluted and spread on 3 different plates according to the microbial species: − P. chrysogenum was cultivated on Rose Bengal Chloramphenicol agar [25], and the plates were incubated at 25 °C for 5 days; − S. epidermidis was cultivated on Chapman agar culture medium, and the plates were incubated at 37 °C for 24 h; − Serratia marcescens was cultivated on nutrient agar culture medium, and the plates were incubated at 30 °C for 48 h.
Finally, bacteria and fungi colonies were counted, and concentrations were expressed in colony-forming units (CFU) per filter surface. The limit of quantification (LQ) was set to 10 CFU/cm 2 .
The possible presence of Zn in the extracted solution, adversely affecting the plating (CFU) results, was evaluated by comparing the microbial concentrations extracted from new and loaded PP and PP/ZPT filters right after contamination, i.e., at t = 0.

Setup for the Study of the Filtration Performance Regarding PM10
The filtration performance of the 2 tested filters was compared regarding PM10 particle collection. The PP/ZPT and the PP filters were loaded with the micronized rice particles at the nominal filtration velocity of 0.12 m/s to evaluate the evolution of their pressure drop and filtration efficiency.
The experimental setup was a straight horizontal filtration device of 5 m length and a section of 9.6 × 9.6 cm. The airflow was ensured by a centrifugal fan located downstream of the filter. The airflow was maintained constant during the test. Two straight lengths of flow stabilization upstream and downstream of the tested filter allowed the generation and measurement of particles using sampling probes. The pressure drop of the tested filter was performed with a differential pressure sensor. Particle counting was ensured by an aerodynamic counter (APS 3321, TSI) from isokinetic sampling upstream and downstream of the tested filter at 5 L/min. The spectral filtration efficiency at time t was quantified from 3 successive measurements upstream/downstream/upstream of the tested filter according to

Setup for the Study of the Filtration Performance Regarding PM10
The filtration performance of the 2 tested filters was compared regarding PM10 particle collection. The PP/ZPT and the PP filters were loaded with the micronized rice particles at the nominal filtration velocity of 0.12 m/s to evaluate the evolution of their pressure drop and filtration efficiency.
The experimental setup was a straight horizontal filtration device of 5 m length and a section of 9.6 × 9.6 cm. The airflow was ensured by a centrifugal fan located downstream of the filter. The airflow was maintained constant during the test. Two straight lengths of flow stabilization upstream and downstream of the tested filter allowed the generation and measurement of particles using sampling probes. The pressure drop of the tested filter was performed with a differential pressure sensor. Particle counting was ensured by an aerodynamic counter (APS 3321, TSI) from isokinetic sampling upstream and downstream of the tested filter at 5 L/min. The spectral filtration efficiency at time t was quantified from 3 successive measurements upstream/downstream/upstream of the tested filter according to where E i is the filtration efficiency of the tested filter at time t of the clogging for particles with diameter d pi ; C down,i is the number concentration of particles with diameter d pi mea-Processes 2021, 9, 1528 7 of 13 sured by the particle counter located downstream of the tested filter; C up,i is the average number concentration of particles with diameter d pi measured by the particle counter located upstream of the tested filter, calculated from the 2 measurements before and after the downstream measurement. The total filtration efficiency at time t was quantified according to The changes in the filter pressure drop and total filtration efficiency during the clogging were expressed as a function of the mass per unit area of collected particles (g/m 2 ) calculated from the mass of particles weighed at the end of the clogging and the particle concentrations measured upstream and downstream of the tested filter.

Water Retention Capacity of the Tested Filters
The water retention capacities of the new and loaded PP/ZPT filters are presented in Figure 3 after conditioning at 50% or 90% RH. The first observation is that the PP/ZPT filters, new or loaded, demonstrate quantifiable water retention capacities. Specifically, the results demonstrate that (i) the level of humidity during conditioning (i.e., 50% vs. 90% RH) influences the water retention capacity of the tested PP/ZPT filters: the water retention capacity of the new and loaded PP/ZPT filters is higher with RH conditioning of 90% than 50% with an increase factor of 38% and 27% and a confidence level of 97.2% and 99.7%, respectively; (ii) the water retention capacity is significantly higher with the loaded PP/ZPT filter than the new PP/ZPT filter by around 18 times for conditioning at 50% or 90% RH (with a confidence level of >99.9%).
New PP filters were also tested after conditioning at 25 • C and 90% RH, and the quantity of water sorbed for a given surface of filtration was slightly lower than that of the new PP/ZPT filters for the same conditions with a confidence level of 97%; by considering the same PP material of the fibers, this result may be explained by the porous structural parameters of the two tested filters.
In conclusion, the presence of water on the new or loaded PP/ZPT filters, required for microbial growth, is confirmed by these results. Moreover, the results indicate that the organic loading of filters, simulated by micronized rice particles under the tested conditions, sorbs a large amount of water, attributed to the nature of the particles used. where is the filtration efficiency of the tested filter at time t of the clogging for particles with diameter ; , is the number concentration of particles with diameter measured by the particle counter located downstream of the tested filter; ̅ , is the average number concentration of particles with diameter measured by the particle counter located upstream of the tested filter, calculated from the 2 measurements before and after the downstream measurement.
The total filtration efficiency at time t was quantified according to The changes in the filter pressure drop and total filtration efficiency during the clogging were expressed as a function of the mass per unit area of collected particles (g/m 2 ) calculated from the mass of particles weighed at the end of the clogging and the particle concentrations measured upstream and downstream of the tested filter.

Water Retention Capacity of the Tested Filters
The water retention capacities of the new and loaded PP/ZPT filters are presented in Figure 3 after conditioning at 50% or 90% RH. The first observation is that the PP/ZPT filters, new or loaded, demonstrate quantifiable water retention capacities. Specifically, the results demonstrate that (i) the level of humidity during conditioning (i.e., 50% vs. 90% RH) influences the water retention capacity of the tested PP/ZPT filters: the water retention capacity of the new and loaded PP/ZPT filters is higher with RH conditioning of 90% than 50% with an increase factor of 38% and 27% and a confidence level of 97.2% and 99.7%, respectively; (ii) the water retention capacity is significantly higher with the loaded PP/ZPT filter than the new PP/ZPT filter by around 18 times for conditioning at 50% or 90% RH (with a confidence level of >99.9%).
New PP filters were also tested after conditioning at 25 °C and 90% RH, and the quantity of water sorbed for a given surface of filtration was slightly lower than that of the new PP/ZPT filters for the same conditions with a confidence level of 97%; by considering the same PP material of the fibers, this result may be explained by the porous structural parameters of the two tested filters.
In conclusion, the presence of water on the new or loaded PP/ZPT filters, required for microbial growth, is confirmed by these results. Moreover, the results indicate that the organic loading of filters, simulated by micronized rice particles under the tested conditions, sorbs a large amount of water, attributed to the nature of the particles used.

Evaluation of the Influence of ZPT on the Filtration Performance Regarding PM10
The evolutions of the filter pressure drop and total filtration efficiency according to the areal mass of particles collected are presented in Figure 4. The results indicate a fairly similar evolution of the increase in the filter pressure drop and total filtration efficiency. The two tested filters lead to similar depositions of the particles during the first in-depth filtration stage and the following stage of cake filtration.
The evolutions of the spectral efficiency of the two tested filters during their clogging are presented in Figure 5. The results indicate that the particle collection efficiency is higher than 96% for the two tested filters whatever the particle diameter and the level of clogging. In particular, for the PP/ZPT filter, for the high level of clogging (i.e., from 36 g/m 2 of areal particle mass collected), we observe a slight decrease in the particle collection efficiency for the submicronic particles, characteristic of electret filters. Basically, according to filtration theory, the filtration efficiency of fibrous filters containing electrostatic charges is enhanced in the beginning of clogging due to the particle collection mechanism by electrostatic forces; the forces disappear with the clogging, leading to a possible decrease in the filtration efficiency if the cake filtration does not compensate it.
To conclude, the two tested filters display a fairly similar filtration performance regarding PM10 particle collection, meaning that the zinc pyrithione has no effect on the filtration performance.

Evaluation of the Influence of ZPT on the Filtration Performance Regarding PM10
The evolutions of the filter pressure drop and total filtration efficiency accord the areal mass of particles collected are presented in Figure 4. The results indicate similar evolution of the increase in the filter pressure drop and total filtration effi The two tested filters lead to similar depositions of the particles during the first in filtration stage and the following stage of cake filtration.
The evolutions of the spectral efficiency of the two tested filters during their cl are presented in Figure 5. The results indicate that the particle collection efficie higher than 96% for the two tested filters whatever the particle diameter and the l clogging. In particular, for the PP/ZPT filter, for the high level of clogging (i.e., fr g/m 2 of areal particle mass collected), we observe a slight decrease in the particle col efficiency for the submicronic particles, characteristic of electret filters. Basically, a ing to filtration theory, the filtration efficiency of fibrous filters containing electr charges is enhanced in the beginning of clogging due to the particle collection mech by electrostatic forces; the forces disappear with the clogging, leading to a possi crease in the filtration efficiency if the cake filtration does not compensate it.
To conclude, the two tested filters display a fairly similar filtration performa garding PM10 particle collection, meaning that the zinc pyrithione has no effect filtration performance.

Evaluation of the Influence of ZPT on the Filtration Performance Regarding PM10
The evolutions of the filter pressure drop and total filtration efficiency according the areal mass of particles collected are presented in Figure 4. The results indicate a fai similar evolution of the increase in the filter pressure drop and total filtration efficien The two tested filters lead to similar depositions of the particles during the first in-dep filtration stage and the following stage of cake filtration.
The evolutions of the spectral efficiency of the two tested filters during their clogg are presented in Figure 5. The results indicate that the particle collection efficiency higher than 96% for the two tested filters whatever the particle diameter and the leve clogging. In particular, for the PP/ZPT filter, for the high level of clogging (i.e., from g/m 2 of areal particle mass collected), we observe a slight decrease in the particle collect efficiency for the submicronic particles, characteristic of electret filters. Basically, acco ing to filtration theory, the filtration efficiency of fibrous filters containing electrosta charges is enhanced in the beginning of clogging due to the particle collection mechani by electrostatic forces; the forces disappear with the clogging, leading to a possible crease in the filtration efficiency if the cake filtration does not compensate it.
To conclude, the two tested filters display a fairly similar filtration performance garding PM10 particle collection, meaning that the zinc pyrithione has no effect on filtration performance.

Cultivable Microbial Concentrations Nebulized Upstream of the Tested Filters during the Tests
The cultivable microbial concentrations (CFU/mL) in the suspension nebulized upstream of the filters during the several tests are indicated in Table 1; the average values were close to 10 11 CFU/mL for S. epidermidis and S. marcescens, and 10 8 CFU/mL for P. chrysogenum. The maximum quantities of species generated upstream of the filters during the 10 min of contamination were calculated assuming a negligible loss of cultivability with generation and a negligible deposit in the setup: around 10 16 CFU for S. epidermidis and S. marcescens, and around 10 12 CFU for P. chrysogenum.

Evaluation of ZPT Transfer during Microorganism Extraction from the Tested Filters
A possible bias in the methodology was investigated regarding the transfer of ZPT in solution during the step of microorganism extraction from the filters and its potential inhibition in the culture-based method.
The microbial concentrations extracted from the new and loaded PP and PP/ZPT filters right after microbial contamination with the microbial suspension are presented in Figure 6. The results indicate that the cultivable microbial concentrations extracted from the tested filters at t = 0 for each species are comparable regardless of the level of clogging of the filters and the presence of ZPT. The cultivable microbial concentrations quantified are around 10 7 CFU/cm 2 for S. epidermidis and S. marcescens, and 10 4 CFU/cm 2 for P. chrysogenum. Thus, the extraction methodology of the filters should not significantly transfer ZPT in the solution because no effect on the cultivability of the tested microorganisms was observed between the PP and PP/ZPT filters.
In addition, the cultivable concentration of the fungi onto the filters right after contamination, i.e., t = 0, was determined with new (N = 7) and loaded (N = 8) filters (regardless of the presence of ZPT in the filters). The average cultivable concentrations (±standard deviations) of P. chrysogenum were not significantly different (9.2 × 10 3 ± 2.7 × 10 3 CFU/cm 2 and 1.5 × 10 4 ± 1.3 × 10 4 CFU/cm 2 for the new and loaded filters, respectively). This result confirms that the quantity of spores of P. chrysogenum coming from the micronized rice loading is negligible compared to that collected during the contamination with the microbial suspension. Processes 2021, 9, x FOR PEER REVIEW 10 of 13 Figure 6. Microbial concentrations extracted from new and loaded PP and PP/ZPT filters right after contamination at t = 0 (averages and ranges).

Antimicrobial Effect of ZPT New Filters
The results of the cultivable microbial concentrations (expressed in CFU/cm 2 ) extracted from the new PP and PP/ZPT filters right after the microbial contamination of the filters (t = 0) and after 8 days of filter conditioning at 25 °C and 50% or 90% RH (t = 8d 50% RH, or t = 8d 90% RH) are presented in Figure 7 for the three microbial species studied.
After 8 days at high humidity (90% RH), the results demonstrate that for the regular PP filter (without antimicrobial treatment), the fungal spore concentration on the filter increases significantly by around 1 log. On the other hand, the Gram-positive bacteria concentration decreases by 2 log, and the Gram-negative concentration decreases by around 0.5 log (with a confidence level of 91% with the t-test). In the presence of antimicrobial treatment in the filtering media, the results show no growth for P. chrysogenum on the new PP/ZPT filter, and the concentration slightly decreases (with a confidence level of 95%). The S. marcescens concentration decreases by 1 log, and the concentration of S. epidermidis decreases by 4 log (i.e., around 90% and 99.99% of inactivation, respectively).
For conditioning at 50% RH, the bacterial survival is limited, with no living bacteria extracted from the new PP and PP/ZPT filters after 8 days, except for a low concentration of S. epidermidis with the PP/ZPT filter, which may be explained by the higher water retention capacity of the PP/ZPT filter; the survival of P. chrysogenum at 50% RH is also limited, with a decrease in the fungal concentration by 2 log for the PP and the PP/ZPT filters.

Antimicrobial Effect of ZPT New Filters
The results of the cultivable microbial concentrations (expressed in CFU/cm 2 ) extracted from the new PP and PP/ZPT filters right after the microbial contamination of the filters (t = 0) and after 8 days of filter conditioning at 25 • C and 50% or 90% RH (t = 8d 50% RH, or t = 8d 90% RH) are presented in Figure 7 for the three microbial species studied.
After 8 days at high humidity (90% RH), the results demonstrate that for the regular PP filter (without antimicrobial treatment), the fungal spore concentration on the filter increases significantly by around 1 log. On the other hand, the Gram-positive bacteria concentration decreases by 2 log, and the Gram-negative concentration decreases by around 0.5 log (with a confidence level of 91% with the t-test). In the presence of antimicrobial treatment in the filtering media, the results show no growth for P. chrysogenum on the new PP/ZPT filter, and the concentration slightly decreases (with a confidence level of 95%). The S. marcescens concentration decreases by 1 log, and the concentration of S. epidermidis decreases by 4 log (i.e., around 90% and 99.99% of inactivation, respectively).
For conditioning at 50% RH, the bacterial survival is limited, with no living bacteria extracted from the new PP and PP/ZPT filters after 8 days, except for a low concentration of S. epidermidis with the PP/ZPT filter, which may be explained by the higher water retention capacity of the PP/ZPT filter; the survival of P. chrysogenum at 50% RH is also limited, with a decrease in the fungal concentration by 2 log for the PP and the PP/ZPT filters.

Antimicrobial Effect of ZPT New Filters
The results of the cultivable microbial concentrations (expressed in CFU/cm 2 ) extracted from the new PP and PP/ZPT filters right after the microbial contamination of the filters (t = 0) and after 8 days of filter conditioning at 25 °C and 50% or 90% RH (t = 8d 50% RH, or t = 8d 90% RH) are presented in Figure 7 for the three microbial species studied.
After 8 days at high humidity (90% RH), the results demonstrate that for the regular PP filter (without antimicrobial treatment), the fungal spore concentration on the filter increases significantly by around 1 log. On the other hand, the Gram-positive bacteria concentration decreases by 2 log, and the Gram-negative concentration decreases by around 0.5 log (with a confidence level of 91% with the t-test). In the presence of antimicrobial treatment in the filtering media, the results show no growth for P. chrysogenum on the new PP/ZPT filter, and the concentration slightly decreases (with a confidence level of 95%). The S. marcescens concentration decreases by 1 log, and the concentration of S. epidermidis decreases by 4 log (i.e., around 90% and 99.99% of inactivation, respectively).
For conditioning at 50% RH, the bacterial survival is limited, with no living bacteria extracted from the new PP and PP/ZPT filters after 8 days, except for a low concentration of S. epidermidis with the PP/ZPT filter, which may be explained by the higher water retention capacity of the PP/ZPT filter; the survival of P. chrysogenum at 50% RH is also limited, with a decrease in the fungal concentration by 2 log for the PP and the PP/ZPT filters.

Loaded Filters
The same experiments were performed as described previously with the filters loaded with micronized rice particles (PM10) containing the fungus P. chrysogenum.
The results presented in Figure 8 demonstrate that the presence of organic particles collected by the filter influences the microbial survival on the tested filters conditioned at 90% RH. The bacterial survival after 10 days is higher for S. epidermidis in comparison to the new filters, particularly for the PP/ZPT filter with antimicrobial treatment, with around a 1.5 log reduction (against 3.5 log with the new filters); for S. marcescens, the concentration decrease is more significant with the loaded filter, probably influenced by the growth of P. chrysogenum on the filters. Indeed, in the presence of organic particles which could be used as a substrate by P. chrysogenum, this strain presents significant growth on the PP filter and also the PP/ZPT filter. The microbial growth on the PP filter was observed by SEM ( Figure 9).
The low relative humidity (50% RH) during filter conditioning has a strong effect on the bacterial survival, with no living Gram-negative S. marcescens extracted from the filters and a sharp decrease in the S. epidermidis concentration by 5-6 log for the two tested filters (i.e., 99.999-99.999% of inactivation). The behavior of P. chrysogenum with the loaded PP and PP/ZPT filters after incubation at 50% RH is almost the same as that with the new filters, with a sharp decrease in the concentration by 1-2 log.
Verdenelli et al. [19] studied the antimicrobial activity of glass fiber HEPA filters treated with phosphated quaternary amine complexes against microbial strains (10 bacteria and 6 fungi). They concluded that the antimicrobial activity is strongly influenced by the working time of the filters, and they observed a decrease in the antimicrobial activity for several strains with used filters compared to unused filters.
The same filters as in the present study were tested over 7 months in realistic conditions with semi-urban outdoor air [26]. The results confirmed the antimicrobial effectiveness of zinc pyrithione regarding fungi cultivated with DRBC agar, i.e., that the fungal concentration on the regular PP filter was significantly higher than that on the ZPT/PP filter, with no influence of the loading of the filter considering a moderate increase in the filter pressure drop by about 30 Pa. No conclusion was established regarding the bacteria because of the high statistical deviation.

Loaded Filters
The same experiments were performed as described previously with the filter loaded with micronized rice particles (PM10) containing the fungus P. chrysogenum.
The results presented in Figure 8 demonstrate that the presence of organic particle collected by the filter influences the microbial survival on the tested filters conditioned a 90% RH. The bacterial survival after 10 days is higher for S. epidermidis in comparison to the new filters, particularly for the PP/ZPT filter with antimicrobial treatment, with around a 1.5 log reduction (against 3.5 log with the new filters); for S. marcescens, the con centration decrease is more significant with the loaded filter, probably influenced by th growth of P. chrysogenum on the filters. Indeed, in the presence of organic particles which could be used as a substrate by P. chrysogenum, this strain presents significant growth on the PP filter and also the PP/ZPT filter. The microbial growth on the PP filter was observed by SEM (Figure 9).
The low relative humidity (50% RH) during filter conditioning has a strong effect on the bacterial survival, with no living Gram-negative S. marcescens extracted from the filter and a sharp decrease in the S. epidermidis concentration by 5-6 log for the two tested filter (i.e., 99.999-99.999% of inactivation). The behavior of P. chrysogenum with the loaded PP and PP/ZPT filters after incubation at 50% RH is almost the same as that with the new filters, with a sharp decrease in the concentration by 1-2 log.
Verdenelli et al. [19] studied the antimicrobial activity of glass fiber HEPA filter treated with phosphated quaternary amine complexes against microbial strains (10 bacte ria and 6 fungi). They concluded that the antimicrobial activity is strongly influenced by the working time of the filters, and they observed a decrease in the antimicrobial activity for several strains with used filters compared to unused filters.
The same filters as in the present study were tested over 7 months in realistic condi tions with semi-urban outdoor air [26]. The results confirmed the antimicrobial effective ness of zinc pyrithione regarding fungi cultivated with DRBC agar, i.e., that the funga concentration on the regular PP filter was significantly higher than that on the ZPT/PP filter, with no influence of the loading of the filter considering a moderate increase in the filter pressure drop by about 30 Pa. No conclusion was established regarding the bacteria because of the high statistical deviation.

Conclusions
Two marketed F7 fibrous filters, one containing zinc pyrithione antimicrobial treatment, were studied. The two filters revealed similar filtration performances during their clogging with PM10 particles, meaning that the antimicrobial treatment did not degrade the filtration performance. The microbial growth on the new and loaded filters contaminated with an airborne consortium of one fungal and two bacterial strains and conditioned at two levels of humidity was studied. At the low humidity value of conditioning (50% RH), with the new or loaded filters, with or without antimicrobial treatment, the microbial population on the filters decreased and possibly did not survive (S. marcescens). At the high humidity value of conditioning (90% RH), the bacteria did not grow on the new filters, and only the fungi were able to develop. The effectiveness of the antimicrobial treatment with zinc pyrithione was confirmed for new filters against the fungus P. chrysogenum. For the loaded filters, the results indicate that the antimicrobial treatment was not more efficient against the fungus P. chrysogenum, the endemic species of the micronized rice particles (PM10) collected by the filters; the two populations of bacteria significantly decreased with or without antimicrobial treatment. To conclude, the data do not support the effectiveness of ZPT in HVAC filters as a single technology implemented for the purpose of protecting occupant health regarding microbial contamination.

Conflicts of Interest:
The authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses or interpretation of data; in the writing of the manuscript, or in the decision to publish the results.

Conclusions
Two marketed F7 fibrous filters, one containing zinc pyrithione antimicrobial treatment, were studied. The two filters revealed similar filtration performances during their clogging with PM10 particles, meaning that the antimicrobial treatment did not degrade the filtration performance. The microbial growth on the new and loaded filters contaminated with an airborne consortium of one fungal and two bacterial strains and conditioned at two levels of humidity was studied. At the low humidity value of conditioning (50% RH), with the new or loaded filters, with or without antimicrobial treatment, the microbial population on the filters decreased and possibly did not survive (S. marcescens). At the high humidity value of conditioning (90% RH), the bacteria did not grow on the new filters, and only the fungi were able to develop. The effectiveness of the antimicrobial treatment with zinc pyrithione was confirmed for new filters against the fungus P. chrysogenum. For the loaded filters, the results indicate that the antimicrobial treatment was not more efficient against the fungus P. chrysogenum, the endemic species of the micronized rice particles (PM10) collected by the filters; the two populations of bacteria significantly decreased with or without antimicrobial treatment. To conclude, the data do not support the effectiveness of ZPT in HVAC filters as a single technology implemented for the purpose of protecting occupant health regarding microbial contamination.

Conflicts of Interest:
The authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses or interpretation of data; in the writing of the manuscript, or in the decision to publish the results.