Enhanced Antimicrobial Activity on Non-Conducting and Conducting Air Filters by Using Air Ions and Grapefruit Seed Extract

The application of grapefruit seed extract (GSE) and air ions to filters was investigated with the goal of collecting indoor bioaerosols. GSE exhibits antimicrobial activity because it contains flavonoids such as naringin. Air ions are usually generated by electric ionizers and can inactivate bacteria, viruses, and fungi. In this study, GSE particles were coated onto three different filters: polypropylene (PP), activated carbon fiber (ACF), and metal foam (MF). Ions were supplied to the filter surface using an ionizer. The rate of surface inactivation of Staphylococcus aureus was measured. When only GSE was applied to the filters, the inactivation rate was about 60%; when air ions were applied, the rate was about 70–80%. When both GSE and ions were applied simultaneously, the inactivation rate increased to 80–90%. The physical collection efficiencies of the as-purchased filters were in the following order: PP > ACF > MF. However, when ions were supplied to the filters, the collection efficiencies increased to over 98% for each filter. The inactivation rate of the ACF filter was the highest (92.5%). Therefore, we conclude that the ACF filter is the most efficient of these three filters for inactivating S. aureus.


INTRODUCTION
Bioaerosols are particulate matter suspended in the air, such as pollen, fungi, viruses, and bacteria.Controlling such aerosols is one of the most important issues for indoor air quality.People are becoming more susceptible to various diseases due to bioaerosols, because airborne diseases are increasing and people are spending more time in crowded spaces.Air-purifying technology and heating, ventilating, and air-conditioning technology are alternatives for reducing this threat.Air filters are usually used to capture pollutants from the air, and would thus accumulate pollutants.Without appropriate action such as replacement or washing, such filters can become another source of bioaerosols because living organisms can grow on their surfaces (Batterman et al., 1995).
To obtain higher inactivation capability using natural products and air ions, essential oil vapors and negative air ions in particular have been used (Tyagi et al., 2010), the combination of which achieved direct inactivation of floating microorganisms.Coating with S. flavescens extract and use of an electrostatic filter (Sim et al., 2015) were also applied to suppress microorganism growth on the filter.The antimicrobial activity increased with the simultaneous application of a coating and an electrostatic force to surfaces.Such synergistic effects are thought to involve antimicrobial activity that occurs on the contact face.The injection of ions can increase the probability of inactivating microorganisms on the filter, because it can increase the likelihood of contact.Moreover, the coating of filters with bactericidal material can reduce bacterial activity from the early stage of a filter's life.During a filter's life, microorganisms accumulate on its surface, which gradually increases the probability of bacteria avoiding direct contact with the antimicrobial material on the filter.To overcome this issue, air ions can be injected into the antimicrobial filter to extend its active life.
In the present study, filters were coated with GSE particles (Woo et al., 2015) and air ions were also prepared (Han et al., 2008) and injected onto the filter surface.Simply immersing the filter media in the GSE solution can cause a significant pressure drop.To avoid this, we used nebulization to deliver GSE material to the filter.To the best of our knowledge, such a combination has not previously been applied to filters.As test bacteria, Staphylococcus aureus was used.S. aureus is Gram-positive and has various negative effects on human health.It is one of the most common species in indoor air (Pastuszka et al., 2000;Wan et al., 2011;Cabo Verde et al., 2015).In this work, we applied a range of different filters to capture airborne S. aureus particles: a polypropylene (PP) filter, an activated carbon fiber (ACF) filter, and a metal foam (MF) filter.PP filters are used in commercial air purifiers, ACF filters are used for deodorizing, while MF filters are used in diesel exhausts due to their ability to withstand high temperatures.PP is a dielectric material, resulting in charged particles easily being able to attach to its surface due to its image force.After a sufficient duration of ion accumulation, incoming ions would be repelled from the PP surface in the opposite direction from the flow of air.Unlike PP filters, ACF and MF filters can accept ions as long as the filter is electrically grounded.Because captured bacteria are located on the filter surface, the inactivation efficiency of the conducting filter would be higher than that of a dielectric filter.
For nebulization, the concentration of bacterial suspension was set to approximately 1 × 10 8 colony-forming units (CFU) mL -1 , as determined by counting the number of CFUs on the agar plate.This solution was prepared by adding two 10-µm loops to 20 mL of sterile water.Sterile water was prepared by autoclaving triple-distilled water for 15 min at 121°C.
Three filters were used in this experiment, as shown in Fig. 1.The melt-blown PP filter (CFX-2HPA; 3AC Ltd., Co., Seoul, South Korea) was of the H13 class, with a fiber diameter of 1.5 µm, and filter thickness of 0.55 mm.The ACF filter was obtained from Modn M-Tech Co., Ltd.(Anyang, South Korea); the fiber diameter was approximately 15 µm, and the thickness of the filter was approximately 3 mm.MF filters (Ni alloy foam filter, N10) were purchased from Aluntum Corp. (Seongnam, South Korea); the nominal cell size was 800 µm, and the thickness was 2.4 mm.These filters were cut to 40 mm by 40 mm for the experiments.

Experimental Procedure
Fig. 2(a) shows the experimental set-up used to make antimicrobial filters, and Fig. 2(b) shows the bacterial aerosol generation set-up for determining the inactivation rate.GSE solution was sprayed with a single-jet Collison nebulizer (BGI Inc., Waltham, MA, USA), supplied with filtered air at 1 L min -1 , and then passed through a diffusion dryer with activated carbon to remove any residual ethanol.The dried natural product particles were deposited onto the test filters for 5 min.The filtration velocity may have effect on coating pattern and the inactivation rate.However, the effect is not evaluated in this paper.We measured the distribution of the sizes of the resulting GSE particles.The most frequently occurring size was 0.6-0.7 µm (Woo et al., 2015).By measuring the weight of the filter, we determined the mass of the GSE particles deposited on the filter to be approximately 2.8 mg.The physical collection efficiency of the non-coated filter was measured using an optical aerosol spectrometer (Grimm Aerosol Technik GmbH & Co. KG, Germany) and the filtration velocity was about 0.01 m s -1 .
S. aureus bioaerosols were generated using the same  kind of nebulizer as in the GSE coating, and passed through a diffusion dryer containing silica gel to remove water vapor (see Fig. 2(b)).We measured the sizes of the bioaerosol particles and found that they most frequently had the same size as the GSE particles (Woo et al., 2015).To determine whether the use of air ions on the test filters had an antimicrobial effect, ions were supplied to the filter surface by a carbon fiber ionizer (Han et al., 2008) with a DC power supply and 1 L min -1 of filtered air during 5 minutes, as shown in Fig. 2(c) after exposing the filter to S. aureus.
After depositing the S. aureus on the filter, we carried out inactivation experiments.The both polarity of ions were tested.This eliminated the possibility of interplay between the collection efficiency and the inactivation rate.To minimize the effect of moisture (Yawootti et al., 2015) in the air, a diffusion dryer with silica gel was also used.The pressure drop of the filters was measured using a digital manometer (Model 350XL/454; Testo, Inc., Sparta Township, NJ, USA).
During the inactivation tests, the generated S. aureus bioaerosols were deposited on the prepared filters for 5 min.The deposited bioaerosols were extracted from the filters in 5 mL of buffer fluid, consisting of phosphate-buffered saline with 0.05% (v/v) Tween 80.The extraction fluid was vortexed for 2 min, agitated in an ultrasonic bath for 10 min, and then vortexed again for 2 min.The extracted fluid was incubated for 24 h at 37°C.To eliminate the antimicrobial effect during the incubation, centrifugation (relative centrifugal force of 5,000 g, 10 min) was carried out.The number of CFUs was counted following incubation.The inactivation rate was defined as follows: where CFU control is the number of colonies per mL recovered from the control filter on which no antimicrobial natural product had been deposited.Between the sets of experiments, the nebulizer was washed and sterilized.

RESULTS AND DISCUSSION
Fig. 3 shows the pressure drops associated with the test filters.The values for the filters were in the following order: PP > ACF > MF.The pressure drop was closely related to the fiber diameter and thickness.The PP filter had a relatively large pressure drop considering its thickness of 0.55 mm.
Filter collection efficiency was measured, as shown in Fig. 4, by assuming that the S. aureus particles were 0.7 µm in diameter (Woo et al., 2015).The efficiency of PP was nearly 100%, while that of ACF was 87% and that of MF was 72%.However, when the negative ions were supplied to the filters, the efficiency increased up to 98% for each filter.This increase in efficiency when ions are supplied to the filters has been reported elsewhere (Shi et al., 2015).The ionizers used in this experiment were located about 20 mm away from the filters.The electrostatic field around the filter caused by a high electric potential (± 10 kV DC @ ionizer) increased the filter collection efficiency, as it acted like an electrostatic precipitator.The charged particles are guided to the electrically grounded filter surface.
Filter quality factors were evaluated as shown in Fig. 5 and are defined as q F = ln(1/(1 -η))/∆P (Hinds, 1999), where η is filter efficiency and ∆P is pressure drop.As the diameters of S. aureus bioaerosols are in the range 0.5-1.0µm, the efficiency η was chosen to be 0.7 µm.The studied filters showed decreasing quality factors with increasing flow velocity.The quality was enhanced when ions were injected into the filters.The quality factors of the ACF and MF filters increased while that of the PP filters remained at similar levels.ACF filters with ions showed the highest factors, except for at flow velocities of 0.01 and 0.04 m s -1 .The ACF filter was not significantly worse than the PP filter.
The ion current was measured by a current meter by changing the number of ionizers and the polarity of supplied DC potential, as shown in Fig. 6.The ion current increased with the number of ionizers and negative polarity of -10 kV was almost double the level of positive polarity (Adachi et al., 1993).This is thought to have been due to the corona discharge mechanism (Huang et al., 2001).The highest value of 12.65 µA can be calculated as 4.74 × 10 18 ions m -3 by using the following equation: where n is the ion concentration, I is the measured current, Q is the flow rate, and e is the elementary charge of 1.60 × 10 -19 C.This value is one of the highest ion concentrations reported (Kim et al., 2011;Panich et al., 2013;Yawootti et al., 2015).The effects of ions on the rate of survival of bacteria on filters are illustrated in Fig. 7. Negative ions were more effective than positive ones in each case, which was attributed to the difference in ion concentration.As the number of ionizers increased, the inactivation rate increased for each filter.The ACF and MF filters showed higher inactivation rates than the PP filter.This is thought to have been due to the difference in the level of ion supply to the filters,  because these filters are made of electrically conductive materials.Ions flow through the bacteria surface as a form of OH -to the filter surface.Sheet resistance was measured using a probe station (Keithley 4200-SCS).The sheet resistance of the PP filter was approximately 1.51 × 10 5 ohm sq -1 , that of the ACF filter was 3.68 ohm sq -1 , and that of the MF filter was 1.95 ohm sq -1 .The electrostatic field strength at the surface of the filter affects the inactivation rate.
When GSE coating was applied to the filters, this tendency was not changed, as shown in Fig. 8.The ACF filter showed the highest inactivation rate among the three filters.Although air ions created by four ionizers were injected into the GSE-coated filters, the above-described tendency remained unchanged.This means that GSE coating process does not change the ion effect significantly.The antimicrobial action seems to occur independently from the experimental results.Single bacteria would be inactivated by the contact with GSE particles or ions.Each probability can be assumed as P GSE and P ion .The total probability of inactivation can be defined as P total = 1 -(1 -P GSE ) × (1 -P ion ).From this model equation, the difference between the results and the model estimation was within 10%.The highest inactivation rate of 92.5% was exhibited by the ACF filter with GSE and ions.The antimicrobial activity arises from the surface.After a sufficient amount of time has passed, the probability of bacteria not being in contact with the filter surface made of antimicrobial material increases.The coverage of GSE on filters will not be 100%, and thus bacterial aerosol can be localized on top of the dust that is captured on the filter.Ion injection can overcome this issue, resulting in a synergistic effect.
As hydrophilic surface properties are important, contact angle measurement (Han et al., 2015) was performed using an SEO 300A (Surface Electro Optics Co., Ltd., Suwon, South Korea), as shown in Fig. 9. Before GSE coating, the contact angles of PP and MF were 123.31° and 85.65°, respectively, while that of the ACF filter was under 10°.After GSE coating, the contact angle of the above filters was also under 10°.Although most flavonoid molecules are hydrophilic, naringin consists of a hydrophilic rhamnose group and a hydrophobic naringenin group and is amphiphilic.A naringin-coated surface will thus also be amphiphilic, enabling it to be applied to various surfaces regardless of its hydrophilicity.In this context, the relative humidity can have an effect on the antimicrobial activity.During the experiment, the relative humidity was maintained at approximately 50%.

CONCLUSION
We have investigated the antimicrobial effect of GSE and ions on air filters of PP, ACF, and MF.Especially for ion injection, filters made of electrically conducting material would be useful to inactivate bacteria on the surface of the filter.The PP filter had the highest collection efficiency among the three filters.Meanwhile, the ACF filter showed the highest inactivation activity for GSE coating and/or ions.When ions were supplied to the filters, the collection efficiency increased.Moreover, the ACF filter is known to have a deodorizing capability.To confer a bactericidal effect on air filters, the ACF filter would be a good alternative compared with the PP filters that are currently widely used for air purifiers.Impregnating the filter fibers with antimicrobial material during the manufacturing process will enhance their performance and reduce the cost.This research should contribute to the development of efficient antimicrobial filter technology.

Fig. 1 .
Fig. 1.Scanning electron micrographs of the filters used: (a) polypropylene (PP) filter (inset: low magnification image of the cross section of the filter), (b) activated carbon fiber (ACF) filter, and (c) metal foam (MF) filter.

Fig. 2 .
Fig. 2. (a) The experimental set-up for (a) grapefruit seed extract coating on filters, (b) Staphylococcus aureus deposition on filters, and (c) air ion injection on filters.

Fig. 6 .Fig. 7 .
Fig. 6.Measured ion current applied to the filters with respect to the number of ionizers.The error bars represent the standard deviation (n = 3).

Fig. 8 .Fig. 9 .
Fig. 8.Comparison of the inactivation rate upon GSE and ion treatments of the filters.The error bars represent the standard deviation (n = 3).