Control of Bioaerosols in Indoor Environment by Filter Coated with Nanosilicate Platelet Supported Silver Nanohybrid ( AgNPs / NSP )

Currently silver nanoparticles (AgNP)-modified filter are widely used to inactivate airborne microbes in indoor environment. However, AgNP is extremely small and thus will penetrate cells membranes to cause cytotoxicity. AgNPs/NSP has been proven to be less cytotoxic to human body. In this study, it was the first time that AgNPs/NSP was used to develop a new antimicrobial air filter with low cytotoxicity. The AgNPs/NSP filter was made by dip-coating of filter with AgNPs/NSP and acrylic resin solution and three different amount of silver on filter were obtained including 12.6, 31.5 and 63 ppm. The filtration efficiency and the antimicrobial activity of AgNP/NSP filter were evaluated by bioaerosols including Escherichia coli and Candida famata in testing chamber and HVAC simulation system under 30% and 70% relative humidity (RH). The results showed that filtration efficiency of AgNPs/NSP-modified filter increased by about 13 to 20% compared to unmodified filter for E. coli but remained almost the same for C. famata. The antimicrobial efficiency of AgNPs/NSP modified filter of 63 ppm was 95.1% for E. coli at RH of 30%. In addition, 91% of antimicrobial efficiency for C. famata was found at RH of 70%. On the other hand, the antimicrobial efficiency of yeast for AgNPs/NSP-modified filter was 97.8% and 86.4% for RH of 30% and 70% respectively when yeast just started to contact with filter in HVAC system. The results suggest that AgNP/NSP-modified air filter can effectively inactivate microorganisms retained on. Therefore, emission of bioaerosols from air filter can be avoided in order to improve the air cleaning technology in indoor environment.


INTRODUCTION
Bioaerosols have been shown to have adverse effects to human health.People with exposure to bioaerosols may get acute pulmonary and respiratory diseases (Peccia et al., 2008).Children under exposure to mold or dampness may increase the risk of developing allergic rhinitis or nonatopic asthma (Jaakkola et al., 2010;Weinmayr et al., 2013).The outbreak of severe acute respiratory syndrome (SARS) in 2003 and H1N1 pandemic in 2009 prompted high attention to bioaerosols.In 2011, Taiwan EPA implemented the Indoor Air Quality Management Act (IAQMA) and started to regulate the concentration of bioaerosols in indoor environment.
Filtration, which can remove particulate matter and bioaerosols simultaneously, is widely used in air purifiers to improve indoor air quality (Lee, 2011).However, microorganisms may grow on air filter and emit harmful volatile organic compounds or obnoxious odor over time (Maus et al., 2001;Moritz et al., 2001).To solve this problem, filter can be treated with antimicrobial agent to inactivate captured microorganisms.For example, natural products such as grapefruit seed extract, sophora extract, and tea tree oil have been used to reduce the number of viable microbes attached on filter (Pyankov et al., 2008;Jung et al., 2011;Han et al., 2015); Yoon et al. (2008) combined silver particles and activated carbon fibers to produce an effective antimicrobial filter; recently, silver-doped TiO 2coated filters are developed to remove Mycobacterium tuberculosis in ambient environment (Thunyasirinon et al., 2015).
Currently silver nanoparticle (AgNP) is one of the most prevalent antimicrobials because it is relatively less toxic and environmentally friendly (Windler et al., 2013) and has therefore been widely used in commercial silver coated filters.However, the silver coated air filters could still potentially release AgNPs into air while using (Quadros and Marr, 2010).Due to its nanoscale size (1-100 nm), AgNPs may penetrate cell membranes and accumulate in cells, which may cause cytotoxicity and genotoxicity (Braydich-Stolle et al., 2005;AshaRani et al., 2009).Therefore, Su et al. (2009) immobilized AgNPs on natural clays (ca.80 × 80 × 1 nm 3 ) that can prevent AgNPs from penetrating into cells, and the nanohybrids still showed excellent antimicrobial activity.
The purpose of this study is to develop a novel antimicrobial air filter to alleviate the accumulation problem of bioaerosols on filter and thus to provide a better indoor environment.Here we developed a filter coated with high surface nanosilicate platelet (NSP)-supported silver nanohybrids but low cytotoxicity.AgNPs do not penetrate into cells due to the intrinsic geometric characteristic of NSP (Su et al., 2011).The filtration efficiency and the survival ratio of microorganisms on filter were examined using a laboratory-scale HVAC (heating, ventilating, and air conditioning) system.

Silver Nanoparticles/Nanosilicate Platelets Solution
AgNPs/NSP solution was synthesized and provided by Professor Jiang-Jen Lin, Institute of Polymer Science and Engineering at National Taiwan University (Li et al., 2010;Su et al., 2011).NSPs (surface area from 80 × 80 to 100 × 100 nm 2 ) were prepared from a Na + -form montmorillonite (Na + -MMT) according to the exfoliation process.30 g of Na + -MMT slurry (1.0 wt% in water) were dispersed in deionized water at 80°C and agitated for several hours followed by the addition of 3.4 g of AgNO 3 solution (1.0 wt% in water).5 mL of methanol was added as reducing agent and the mixture was stirred at 600 rpm for 3 hours until the color changed from yellow to dark reddish brown, indicating the reduction of Ag + to Ag 0 .The reported number average mean diameter of AgNPs was 30 nm (Dong et al., 2009).

Filter Modification
Antimicrobial filters were obtained by dip-coating method.Non-electrostatic melt-blown polypropylene filters (Mytrex Industries, Inc., Taiwan) weighting 26 g m -2 with a thickness of 0.26 mm were employed in this study.The filtration efficiency was greater than 90% when tested against 0.3 µm of NaCl particles at a flow rate of 32 LPM and the pressure drop was less than 0.3 mm H 2 O at a face velocity of 5.33 cm s -1 .The filters were immersed in the solution comprised 10% v/v of acrylic resin emulsion (First Chemical Corp., Taiwan) and 0.1, 0.05, and 0.02 wt% of AgNPs/NSP, respectively.After fully soaked with coating solution, the filters were dried at 55°C for 24 hours and sterilized by exposing to ultraviolet radiation (254 nm) for 1 hour.The morphology of the filter was observed by scanning electron microscopy (SEM, JSM-6510LV, JEOL, Japan) and the concentrations of silver nanoparticles were measured by atomic absorption spectrometer (AAnalyst 800, PerkinElmer Inc., USA) in accordance with the standard method NIEA A307.10C of Taiwan EPA.In summary, the filter was immersed in HCl/HNO 3 acid solution and heated for 30 minutes to extract inorganic matter.The final extraction solution was then ready for the quantification of silver nanoparticles by was atomic absorption spectrometer.The concentration of silver on filter for three coating solution of AgNPs/NSP was 12.6, 31.5, and 63 ppm respectively according to the measurements from atomic absorption spectrometer.

Microbial Suspension
The microbial suspension was prepared for bioaerosols generation.Escherichia coli (BCRC 14894) and Candida famata (BCRC 21681) were selected as test microorganisms in this study.E. coli is a rod-shaped Gram-negative bacterium commonly used for bioaerosols research (Lee et al., 2008;Schwarzmeier et al., 2013;Xu et al., 2013), whereas C. famata is a frequently encountered airborne fungus in Taiwan that has been selected as a challenging bioaerosol species (Lin and Li, 2002;Yu et al., 2008).To obtain bioaerosols, a loopful of E. coli was inoculated into 40 mL of Luria-Bertani broth (Difco) in a 250 mL Erlenmeyer flask and incubated on an orbital shaker at 150 rpm and 37°C for 24 hours.C. famata was inoculated into 40 mL of Yeast Mold broth (Difco) and incubated at 150 rpm and 25°C for 48 hours.The suspension was centrifuged at 3,200 rpm for 10 minutes and the supernatant was discarded.The residue was resuspended in 0.1% peptone water (LP007 bacteriological peptone, Oxoid Ltd.) and the final concentration was 10 8 -10 9 CFU mL -1 .

Sampling and Analysis of Bioaerosols
BioSampler (SKC Inc., USA), were employed for bioaerosols sampling.For BioSampler, the sampling flow rate was 12.5 LPM and the collection liquid was 20 mL of 0.5% peptone water (Oxoid).After sampling, the collection liquid were serially diluted and plated on tryptic soy agar (TSA, Difco) for E. coli and malt extract agar (MEA, Difco) for C. famata.TSA plates were incubated at 37°C for 24 hours and MEA plates were incubated at 25°C for 48 hours.The concentration of bioaerosols obtained by BioSampler (CFU m -3 ) is calculated according to Eq. ( 1): where f is the dilution factor (dimensionless), N is the counts of microbial colony (CFU), v s is the volume of total collection liquid (mL), v is the liquid volume for plating (mL), Q s is the sampling flow rate (LPM), and T is the sampling time (min).

Filtration Test
All the air entered the system was pretreated with a HEPA filter (12144 HEPA capsule filter, PALL Corp., USA) and a diffusion dryer to remove particulate matter and water vapor, respectively.The air stream were separated into three streams and the air flow rates were controlled by mass flow controllers.One of the air streams was used for bioaerosol generation at flow rate of 2 LPM.The relative humidity was controlled by allocating the amount of air streams passing through gas washing bottle (with sterilized distilled water inside) and through diffusion dryer.The total air flow rate of the experimental system was 12.5 LPM.The air flow and relative humidity were calibrated and adjusted using Gilibrator-2 primary air flow calibrator (Gilian Instrument Corp., USA) and a thermohygrometer (HP21, Rotronic Instrument Corp., USA), respectively.Test filter was installed on a stainless steel holder of 5.4 cm in diameter inside the experimental chamber.Bioaerosols was generated by atomizing microbial suspension using Collison three-jet nebulizer (BGI Inc., USA).The bioaerosols then passed through a silica gel diffusion dryer to remove water vapor followed by a Kr-85 aerosol neutralizer (model 3077, TSI Inc., USA) to obtain Boltzmann charge equilibrium before entering the experimental chamber.The outlet air was sampled by BioSampler.The overall experimental system is illustrated in Fig. 1.
The filtration efficiency of filter was determined by the following steps.To avoid fluctuation, the bioaerosol generation unit and the relative humidity control unit were stabilized for 20 minutes.The background concentration of bioaerosols was measured with sampling time of 10 minutes.The filter was then installed on the filter holder and the concentrations of bioaerosols were measured for 40 minutes (10 minutes per sampling × 4 sampling).The face velocity was controlled to be 8.46 cm s -1 and the experiments were conducted at RH 30% and RH 70%.The experiment was conducted in duplicate.The filtration efficiency (η f ) is calculated according to Eq. ( 2): 1 100 where C f,t is the concentration (CFU m -3 ) of bioaerosols with filter at t minutes and C is the concentration (CFU m -3 ) without filter.
In addition, the microbial activity on filter over time was evaluated by comparing the survival ratio (S t ).We defined that the survival ratio is the amount of culturable microorganisms on filter at a specific time to that at initial time, as calculated by Eq. ( 3): where N e,o is the amount (CFU) of microorganism obtained from filter immediately by extraction process, N e,t is the amount (CFU) of microorganism obtained from filter after t minutes by extraction process.To obtain the CFUs, the tested filters were immediately dismounted from holder and placed into a stainless steel chamber at RH 30% or 70% for 0 min, 10 min, 100 min, and 1000 min.The filters were then put into a 50 mL conical tube with 30 mL of extraction solution consisted of 0.1% peptone (Oxoid) and 0.01% tween 80 (Nacalai tesque), and vortexed vigorously for 2 minutes followed by ultrasonic agitation for 10 minutes.The CFUs of the final extraction solutions were obtained by serial dilution and spread plate method.The experiment was conducted in duplicate.

HVAC System Test
To simulate the bioaerosols in indoor environment, a simplified HVAC system comprised an 80 × 80 × 80 cm 3 stainless steel chamber, a 6 × 6 cm 2 ventilation duct, a 3.8 × 3.8 cm 2 filter holder, a mixing fan, and a return fan, was used in this study (Fig. 2).The ratio of total flow rate (Q t ) to chamber volume (∀) was 1.5 h -1 and the recirculation rate was 50%.Therefore, the ratio of inlet air flow rate (Q in ) to chamber volume was 0.75 h -1 and the ratio of recirculation air flow rate (Q re ) to chamber volume was 0.75 h -1 .To start with, the microbe-laden air went through the chamber at flow rate of 3 LPM for 90 min.Then the return fan was switched on and BioSampler was used for sampling at t = 0, 20, 40, 60, and 80 min.The sampling time was 5 min for each sampling.We defined that the reduction ratio (E t ) is the amount of culturable microorganisms in HVAC system at a specific time to that at initial time, as calculated by Eq. ( 4): where C 0 is the bioaerosol concentration (CFU m -3 ) in HVAC system at initial time, and C t is the bioaerosol concentration (CFU m -3 ) in HVAC system after t minutes.
The experiments were conducted in duplicate and the results were the mean values of two measurements.
The steps to obtain the CFU on tested filter was the same as described in Filtration Test Section, except that the filters were placed into a stainless steel chamber for 0 hour, 12 hours, and 24 hours.The experiment was conducted in duplicate.

Statistical Analysis
Data were analyzed by using one-way ANOVA with significance level α = 0.05.A p ≤ 0.05 was regarded as significant.If significant differences were observed, Tukey post-hoc test was used to compare each set of experimental data.

Characterization of the AgNPs/NSP-Modified Filter
Fig. 3 shows the morphology of AgNPs/NSP-modified filter using SEM.The white dots deposited on the fiber was about 100 × 100 nm 2 , which was consistent with the size of NSP, indicating that AgNPs/NSP was successfully coated on the filter.The total amount of silver nanoparticles on filters was quantified since it is one of the factors associated with antimicrobial activity.The amount of silver coated on filter was 216 mg m -2 , 112 mg m -2 , and 42 mg m -2 by immersing the filter in 0.1, 0.05, and 0.02 wt% of AgNPs/NSP solution, which was approximate 8.31 g,  4.31 g, and 1.62 g per kilogram of filter.The minimum application rate of AgNPs was recommended in a range of 10 to 100 mg kg -1 textile (Windler et al., 2013).

Filtration Test of the AgNPs/NSP-Modified Filter
The filtration efficiency for E. coli.was depicted in Figs.4(a) and 4(b).The filtration efficiency of un-modified filter under RH 30% and RH 70% was, on average, 69.7 ± 6.8% and 72.2 ± 3.9%, respectively, in a 40-minute filtration process.The efficiency of AgNPs/NSP-modified filter, however, increased by 13.7 to 19.7% at relative humidity of 30% and 14 to 19.9% at relative humidity of 70%.Statistically significant difference between groups was determined by one-way ANOVA under RH 30% (F (3, 12) = 15.19,p < 0.001) and RH 70% (F (3, 12) = 30.950,p < 0.001).Post hoc analysis revealed that the filtration efficiency of un-modified filter was significantly lower than that coated with 12.6 ppm (p = 0.002), 31.5 ppm (p = 0.001), and 63 ppm (p < 0.001) of AgNPs.These results indicate that the modification process may strengthen the capability of filter to capture particulate matters.One possible reason of this phenomenon is that the filter structure changed during modification process.The overall efficiency (diffusion and interception) is related to many parameters including the fiber diameter and solidity fraction of filter (Lee and Liu, 1982).
The filtration efficiency for C. famata was illustrated in Figs.4(c) and 4(d).No significant differences were observed between unmodified and modified filters with AgNPs/NSP of three silver concentrations under RH 30% (F (3, 12) = 1.92, p = 0.180) and RH 70% (F (3, 12) = 0.21, p = 0.888).The results indicated that modification process did not affect the filtration.It is also noteworthy that filtration efficiency of E. coli (82.8 ± 8.9%) at RH of 30% was significantly lower than that of C. famata (91.7 ± 3.4%), t(19.25)= -3.740,p = 0.001.Similarly, filtration efficiency of E. coli (85.0 ± 8.4%) at RH of 70% was also significantly lower than that of C. famata (96.3 ± 1.3%), t(15.71)= -5.311,p < 0.001.This may be due to the geometric mean diameter (GMD) of microorganisms.In general, the filtration mechanism is dominated by diffusion if particle diameter is smaller than 0.1 µm, and dominated by impaction and interception if particle diameter is larger than 1 µm.When the particle diameter is between 0.1 µm and 1 µm, the particle is almost free to the aforementioned mechanisms, and thus will have lowest filtration efficiency.Because the GMD of E. coli is about 0.78 µm, and the GMD of C. famata is about 2.44 µm, the filtration efficiency for E. coli will be lower than that for C. famata.
The comparison of the survival ratio between E. coli and C. famata under different relative humidity was made and depicted in Fig. 5.The survival ratios of C. famata were found to be higher than those of E. coli for unmodified and modified filters.The unmodified filter was the control group and death of microorganisms on unmodified filter was due to natural decay.It is noteworthy that modified filers coated with 12.6, 31.5 and 63 ppm of AgNPs/NSP showed lower survival ratio than that of unmodified filter for both E. coli and C. famata.In other words, AgNPs/NSP modified filters showed antimicrobial effect compared to unmodified filters.The survival ratio of C. famata on un-modified filter at 10 min exceeded 100% and still possessed more than 50% the survival ratio regardless of the relative humidity.It indicated that fungi might grow on unmodified air filter.However, average survival ratios of E. coli for 12.6, 31.5 and 63 ppm of AgNPs/NSP-modified filter were decreased by 44%, 86% and 95% compared to that of unmodified filter at 10 minutes after E. coli bioaerosols were captured under relative humidity of 30%.At relative humidity of 70%, average survival ratio of E. coli was decreased by 4.6%, 54.5% and 61.3% for 12.6, 31.5 and 63 ppm of AgNPs/NSPmodified filter compared to that of unmodified filter at 10 minutes after E. coli bioaerosols were captured.The average survival ratios of C. famata for 12.6, 31.5 and 63 ppm of AgNPs/NSP-modified filter were decreased by 14.9%, 7.7%, and 5.8% compared to that of unmodified filter at 10 minutes after C. famata bioaerosols were captured.At relative humidity of 70%, the average survival ratios of C. famata for 12, 31.5 and 63 ppm ppm of AgNPs/NSP-modified filter were decreased by 53.5%, 70.2%, and 90.8% compared to that of unmodified filter at 10 minutes after C. famata bioaerosols were captured.At relative humidity of 30% and 10 minutes after E. coli were captured on filter, the antimicrobial efficiency of modified filter of 63 ppm silver concentration was 95%.In addition, the antimicrobial efficiency for C. famata was 90.8 % for modified filter of 63 ppm silver concentration at relative humidity of 70% and 10 minutes.Therefore, AgNPs/NSP-modified filter demonstrated the antimicrobial effect on E. coli and C. famata bioaerosols and the effect was significant except for the C. famata at reative humidity of 30%.This may be associated with higher resistance of C. famata to AgNPs/NSP than that of E. coli so that fungi have better detoxification system than bacteria (Panacek et al., 2009).Besides, the release mechanism of AgNPs (Xiu et al., 2011) may also affect the biocidal activity, as shown in Eqs. ( 5)-( 6): Since the silver ions release from AgNPs involves the oxidation of zero-valent silver and proton transfer, the protons in water vapor may help release silver ions, and thus enhances the antimicrobial activity of AgNPs/NSP-modified filter under high relative humidity.Another factor to affect the antimicrobial efficiency was the particle shielding effect on air filter.It was reported that microbials may continuously aggregate on filter and thus prevent them from being inactivated (Lee et al., 2009).Therefore, further study is needed to elucidate the shielding effect on antimicrobial efficiency.

Microbial Reduction Ratio in Simulated HVAC System
The reduction ratio of bioaerosols in simulated HVAC system was depicted in Fig. 6.Figs. 6(a) and 6(b) show that around 70% of the total culturable E. coli rapidly decreased within 20 min and less than 5% of culturable E. coli were detected within 80 min whether at RH 30% or RH 70%.No statistically significant differences were found between natural attenuation (without filter) (p = 0.852), pristine filter (p = 0.454), and modified filter (p = 0.484).These results may be due to that E. coli is a sensitive species and not a common species in ambient air.Therefore, it would rapidly become inactivated when aerosolized into air or attached on filter.Similar results have been reported by previous study (Jung et al., 2011).
Meanwhile, the natural attenuation of yeast was more rapid at RH 30% (Fig. 6(c)) than that at RH 70% (Fig. 6(d)).It was mainly related to the desiccation of microbes at low humidity (Pigeot-Remy et al., 2014).Moreover, half of the total culturable C. famata remained in the HVAC system even after an hour.A high standard deviation of total culturable species appeared at the first 20 min at RH 70% for the filtration of un-modified filter, which might be due to the effect of humidity.Although the difference between microbial reduction ratio of natural attenuation and that with filter was not obvious at low relative humidity but it was higher at high humidity when the system was equipped with filter.The reason may be that more microorganisms were captured by the filter at high humidity, and weaker desiccation effect.Vandenbrouckegrauls et al. (1995) reported that the size of the particles retained by the filter is determined by the size of the droplets under high humidity condition, while it is determined by the size of the microorganism under low humidity condition.
In order to further confirm the results of the microbial reduction ratio, the activity of microorganisms on filter in HVAC system were also examined and the results are shown in Fig. 7.It was found that the amount of E. coli captured by the filter was scarce even under 70% humidity (Fig. 7(a)) and no significant difference was observed between pristine filter and modified filter under RH30% (p = 0.499) and RH70% (p = 0.374).Both microbial activity and reduction ratio results indicate that E. coli is not an appropriate species for prolonged (24 hours in our case) bioaerosols experiments because of its sensitivity to ambient environments.Besides, E. coli will enter VBNC (viable but non-culturable) state under stressed environments, which may mislead researchers to count microorganisms as dead state and thus results in experimental bias.results were found in modified filter (30.8 ± 8.9 CFU) and un-modified filter (226.6 ± 40.2 CFU), t(4) = 8.224, p = 0.001.Therefore, the antimicrobial efficiency of yeast for AgNPs/NSP-modified filter was 97.8% and 86.4% at relative humidity 0f 30% and 70% respectively when yeast just started to contact with AgNPs/NSP in filter.On the other hand, the amount of culturable C. famata on unmodified filter after 24 hours was higher than that after 12 hours at RH 70%, suggesting that microorganisms may even grow on filter if the conditions is favorable for them and thus will become an emission source of indoor bioaerosols (Batterman and Burge, 1995).Previous study has shown that the molds and dampness were associated with sick building syndrome (Zhang et al., 2012).It is noteworthy that the amount of culturable C. famata on AgNPS/NSP modified filter remained almost the same over 24 hours.In other words, the antimicrobial effect of AgNPS/NSP modified filter can inactivate C. famata captured to control the problem of growth of microorganisms.Because filtration is a common method for indoor air cleaning, AgNPS/NSP can be used to control the growth and accumulation of microorganisms.Therefore, efficiency of inactivation bioaerosols will be enhanced and the emission of microorganisms from filter can be avoided.

CONCLUSIONS
In this study, a AgNPs/NSP-modified air filter was made and the characteristics including the filtration efficiency and the antimicrobial efficiency were assessed.The filtration efficiency for bacterial aerosols was found to increase for Escherichia coli.After modification process but remained almost the same for Candida famata.The antimicrobial effects of AgNPs/NSP modified filter were found according to the survival ratio of Escherichia coli and Candida famata captured on filter of filteration test and HVAC system test.The antimicrobial efficiency of AgNPs/NSP modified filter for Escherichia coli at relative humidity of 30% was higher than that of relative humidity of 70%.On the contrary, the antimicrobial efficiency of AgNPs/NSP modified filter for Candida famata at relative humidity of 70% was higher than that of relative humidity of 30%.The antimicrobial efficiency was as high as 95.1% for Escherichia coli at relative humidity of 30% and 10 minutes after Escherichia coli were captured on filter.In addition, 91% of antimicrobial efficiency for Candida famata was found at relative humidity of 70%.The results suggest that AgNP/NSP-modified air filter can be applied to air cleaning and it would not only remove bioaerosols but also inactivate them for protection of human health in indoor environment.

Fig. 1 .
Fig. 1.Experimental setup of filtration test system.The arrow indicates the direction of the air flow.

Fig. 3 .
Fig. 3. SEM observation of AgNPs/NSP-modified filter.The white dots on the fiber is possibly the AgNPs/NSP.