Fabrication of nanofiber filters for electret air conditioning filter via a multi-needle electrospinning

With the rapid development of human society, a large amount of the fine suspended particulate matter (PM) emitted by human activities (especially, manufacturing and transportation), of which PM2.5 (PM with aerodynamic diameter da ≤ 2.5 μm) caused severe air pollution, has raised concerns about health.^1–4 These PM come in the form of nanoaerosols and nanoparticles, which are usually hundreds of nanometers and less, presenting in high concentrations from pollutants to viruses. For example, the recent novel coronavirus that has infected many people, as well as the previous SARS and MERS viruses, is also of size ∼100 nm.⁵ Therefore, effective protective measures are urgently required to protect people from PM, and filtration is one of the main methods used for removing fine particles from air. From the melt-blown method of traditional filter preparation, the fibers are typically made of microfibers 2 μm–20 μm in size. However, they are not so effective in filtering nano-pollutants because of their large fiber diameter. Of the available filters, fortunately, fibrous filters are proven relatively efficient and economically applicable in removing micro-aerosols, submicrometer aerosols, and nanoaerosols (less than 100 nm) and are widely applied in the areas of respirators, indoor air purification, vehicle air filtration, industrial gas filters, and water treatment.^6–10 

Attributed to the electrical attraction between fibers and particles with net or induced charge, electret fibrous filters have been of increasing interest for researchers recently.^6,11–17 Electrostatic forces from surface/volume charges or dipoles open up possibilities for low air resistance, high-filtration, and thinner filter preparation.¹⁸ To exhibit electret properties, materials (usually polymers) should have sufficient charge traps or be easily polarized.^11,19 To date, several polymers, including polypropylene, polyurethane, and polyvinylidene fluoride (PVDF),²⁰ have been used to make fibrous filters and are studied for their chargeability and filtration performance.^16,21

According to reports, Sim et al. observed that the polyurethane fiber filter treated by corona discharge had about 30% improvement in filtration efficiency up to 80.9% for KCl particles ranging from 20 nm to 660 nm, compared with the uncharged filter.²¹ Sun et al. achieved the preparation of the multi-layer PVDF electret filters with high strength, low air resistance, and high filtration, which bring a great performance in aerosol removal for long-term aerosol filtration.^5,22 Based on the available research, electret fiber filters are considered to be the subject of a new generation of filters with the advantages of low air resistance and high filtration. However, the industrial production of electret fiber filters is still in an immature stage, lacking the research on this aspect.

Due to their outstanding physical properties, such as good electrical insulation, chemical resistance, ferroelectricity, thermal properties, and biocompatibility, PVDF membranes have been extensively investigated in the fields of sensors, liquid handling, distillation, and gas absorption.^23–26 In addition, the properties of high hydrophobicity and extremely low conductivity may endow PVDF filters with good charge stability and stable filtration performance.^27,28 Moreover, under humid conditions where bacterial growth might pose a risk, the use of silver nanoparticles should be very effective against bacterial growth due to the transfer of silver ions and whole nanoparticles, which can disrupt the vital functions of the bacterial cells. Currently, the antimicrobial effect of silver is exploited in a very diverse set of applications ranging from simple consumer goods to complex medical devices. In addition, how and in what form silver is introduced in these applications varies widely.^29–31 Li et al. reported the one-step preparation of ultrafine poly(acrylonitrile) fibers containing silver nanoparticles, thereby using DMF as the reducing medium of silver nitrate that successfully reduced the Ag⁺ ions.³¹ This method greatly simplifies the preparation of the antibacterial nanofiber solution. Meanwhile, the silver-loaded nanofibers show slightly higher activity than nanoparticles.³² Therefore, the addition of silver nitrate to the spinning solution can theoretically provide antibacterial properties to the fiber. However, high crystallinity of PVDF generally impedes conductivity and pure PVDF with varying beading structures. Bingbing et al. successfully prepared polyvinylpyrrolidone (PVP)/PVDF composite micro/nanofibers with good fiber morphology.³³ Polyvinylidene fluoride and polyvinylpyrrolidone polymers incorporated with carbon black nanoparticles (50 nm) were electrospun to functional fibers by Jabbarnia et al.³⁴ Therefore, PVP was added in the spinning solution in order to account for conductivity limitation and beading structures, increasing electrospinnability of the spinning solution. Although some researchers have studied the electrospinning preparation of PVDF fibers,^35–38 only a few investigated the preparation of PVDF fiber filter membranes in multi-needle electrospinning.

In our research, we combined PVDF fibers with air conditioning filter substrates to make nano fiber air conditioning filters, and the effects of relevant parameters in multi-needle spinning on their fiber morphology, deposition law, and filtration performance were investigated. By combining the advantages of PVDF and traditional air conditioning filters, adding antibacterial silver ions, we managed to fabricate electret air conditioning filters with high filtration efficiency, low pressure drop, and antibacterial performance.

Polyvinylidene fluoride (PVDF) with a molecular weight (MW) of 530 000 and polyvinylpyrrolidone (PVP) with a molecular weight (MW) of 1 300 000 were purchased from Sigma-Aldrich (USA). Silver nitrate (AgNO₃) and N,N-dimethylformamide (DMF) were obtained from Aladdin Chemical Co. All reagents were of analytical grade and used as received.

In a typical preparation, the PVDF solution was prepared by dissolving 18 wt. % PVDF pellets into DMF at 70 °C for 24 h. Then, we increase the viscosity of the spinning solution by adding PVP to the PVDF solution. Meanwhile, we add 1wt. % silver nitrate crystals as an antibacterial material additive. The spinning solution material parameters are tested at 30 °C, as shown in Table I. The needle-based electrospinning machine (Model MF-001, Foshan Lepton Precision Measurement And Control Tech Co. Ltd.) was employed to prepare PVDF nanofiber mats. As shown in Fig. 1, the electrospinning apparatus consists of a syringe pump, a grounded collector, and a high voltage supply set. The gear pump was used to feed the polymer solution into a 20-gauge steel needle tip (ID = 0.6 mm) at the rate of 60 ml h⁻¹. The electrospinning process was carried out with a voltage of 40 kV/45 kV/50 kV/55 kV/60 kV and a tip-to-collector distance (traveling distance) of 30 cm. The temperature was kept at 25 ± 1 °C, and the relative humidity (RH) was controlled at 30%/40%/50%/60%/70%, respectively. As solution oozed out of the steel needle tip, under the high electrical field, the conductive liquid formed a Taylor cone. When the electrical force overcame the interfacial tension of the liquid, a thin fiber jetted toward the ground electrode. The diameter of the fiber reduced continuously due to the repulsion of the positive charges deposited along the fiber together with the evaporation of the solvent. The fiber diameter became ultimately less than 1 μm. The fibers were deposited on the surface of a collector covered by the substrate mesh or aluminum foil. After spinning, the PVDF nanofiber filters (coated on substrate net or aluminum foil) were then dried in a vacuum oven at 45 °C overnight to remove the residual solvent. To get single-layer nanofiber mats with different basis weights (W), different production rates were used, as the amount of fiber deposition is proportional to the electro-spinning (hereafter, simply referred to as spinning) time. In this case, it was reasonable to assume that the filter thickness (Z) increased with the spinning time of the single nanofiber layer. To a lesser extent, the fiber packing density (α) may also increase slightly with spinning time when W is large; otherwise, for smaller W, it can be assumed to be reasonably constant producing a constant porosity mat. The other configuration of filters applied in our research was made by stacking up a different number of nanofiber filters by different production rates to form a multi-layer filter.

The surface morphology of nanofiber filters was analyzed by scanning electron microscopy (Model TM303, Hitachi, Tokyo, Japan). The fiber mean diameters were measured by the software ImageJ (NIH, USA).

The spinning solution in Table I was tested in the spinning equipment. The liquid supply rate was set to 60 ml/h, the voltage was set to 60 kV, and the aluminum foil was used for receiving. Controlling the environmental humidity to 30%, the observation of the fiber morphology using an electron microscope is shown in Figs. 2(a) and 2(b) that compares electrospun pure PVDF fibers and PVDF/PVP fibers. According to Fig. 2, the addition of PVP helps account for reducing beading structures. Then, the spinning effect was tested at an ambient humidity of 30%–70% RH. The observation of the fiber morphology using an electron microscope is shown in Figs. 3(a) and 3(b). From Figs. 3(a) and 3(b), it can be seen that after the viscosity is increased by the PVP material, even if at a relative humidity of 70% RH, the solution is only agglomerated by some undispersed materials, without the beading structure. According to Fig. 4(c), we know that the fiber diameter of most fibers was distributed between 200 nm and 400 nm in the 30%RH environment, and the fiber diameter distribution becomes wider with increasing humidity. Figure 4 shows the statistics of fiber diameter distribution under different humidities. It can be seen that as the humidity increases, the fiber diameter increases. At 30% RH, the statistical Coefficient of Variation (CV) of the fiber diameter is the smallest, indicating that the fiber filament diameter is the most uniform when the ambient humidity is 30% RH under this production condition. Therefore, at an ambient humidity of 30% RH, the fibers are thin and uniform.

The traditional air conditioning filter is made of ordinary polymer mesh without filtering performance, which can only intercept some large particles at the air inlet or air outlet of the air conditioner. Therefore, it cannot filter PM2.5 and other inhalable particles in the room. Nanofibers have unique performance advantages for filtering small particles such as PM2.5, but nanofibers have little self-supporting performance. Therefore, we can combine nanofibers and traditional air conditioning filters by using electrostatic spinning technology to spray nanofibers on traditional air conditioning filters, achieving the performance of PM2.5 dominated by nanofibers, and finishing high-efficiency, low air resistance multifunctional nanofiber air conditioning filters, as shown in Fig. 5.

The substrate meshes used in this experiment include polyester meshes, the number of polyester meshes is 40, 60, 80; nylon meshes, the number of nylon meshes is 80, 120; and polyamide meshes, the number of polyamide meshes is 100, 120. We set the ambient temperature to 30 °C, the liquid supply rate to 60 ml/h, the reciprocating speed of the nozzle to 150 mm/s, and the reciprocating stroke to 120 mm. Both the original filtration efficiency and air resistance of the sieve holes of the substrate are zero. The TSI-8130 device was subsequently used to test the filtration performance of the substrate meshes with the nanofiber for particles with a particle size of 0.3 µm at a flow rate of 32 l/min. The schematic diagram of this experiment is shown in Fig. 6, which realizes multi-nozzle joint work. Figure 9(d) is a comparison of the quality factor (QF) of different substrate filter membranes at different production rates. At 4 mm/s, the maximum QF value is 27.07 for polyamide 100 meshes. In addition, at 8 mm/s, the data group that does not reach high efficiency filtering is excluded, so the highest QF value is 24.05 for nylon 80 meshes. However, the highest efficiency membrane at 16 mm/s is 26.08 for nylon 120 meshes.

First, the spinning results of different receiving substrate meshes at a winding speed of 8 mm/s and different voltages were investigated. The spinning results are shown in Figs. 7(a)–7(d). Among them, Fig. 7(a) is a comparison chart of the average fiber diameter and the deviation of the fiber diameter of the polyester mesh with different meshes under different voltages. It can be seen that as the voltage increases, the fiber diameter decreases, and the smaller fiber diameter is beneficial to improve the filtration efficiency. Meanwhile, compared with different materials, the fiber diameter is also different, which is mainly because the effect of different mesh sizes on the space electric field is different. From Fig. 7(b), we can see that as the spinning voltage increases, the filtration efficiency of the substrate meshes with the nanofiber becomes higher. This is mainly because as the voltage increases, the fiber diameter becomes finer. As a result, the fiber layer is denser and the fiber coverage is more uniform. According to the data, the polyester 80 mesh substrate has always maintained a high filtration efficiency, which is above 80%, while the polyester 60 mesh substrate can reach a filtration efficiency of 80% or more at a voltage of 55 kV or more. However, for the polyester 40 mesh substrate, the filtration efficiency is less than 60%. As shown in Fig. 7(c), we compared the air resistance of different meshes under different voltages, learning that as the voltage becomes larger, the air resistance also increases. Then, in Fig. 7(d), we compared the filter membrane QF. Although the polyester 40 mesh with the nanofiber has a higher QF value than the other two materials, its filtration efficiency is less than 80%, without a high efficiency filtration function. Therefore, it is not discussed. In addition, the polyester 60 mesh only discusses the case of 55 kV and 60 kV. As shown in Fig. 7(d), it can be seen that the polyester 60 mesh material has a larger QF value at 55 kV, which is mainly due to the small air resistance. Compared with the polyester 80 mesh material, the QF value is the largest at 40 kV and is the smallest at 45 kV.

The SEM images of the fiber deposition effect under different voltages are shown in Fig. 8. It can be seen that as the voltage increases, the denser the fibers deposited on the substrate meshes, and the data comparison results in Fig. 8 are verified. Comparing the SEM images of polyester meshes with different meshes number, the uniformity of fiber deposition of polyester 40 mesh is very poor, and it is basically only deposited in the middle part of the mesh area. Therefore, the filtration efficiency and air resistance of polyester 40 mesh are worse. The fiber deposition on polyester 80 mesh is the most uniform. This is mainly because the smaller the mesh, the weaker the electric field interference in the middle area of the mesh so that the electric field is more uniform, and it is easier to make the nanofibers uniformly deposited, which can guarantee a relatively high filtration efficiency. Research on different voltages shows that voltage can obtain finer fibers and form a dense interception layer, so higher filtration efficiency can be obtained. When the production rate is 16 mm/s, only polyester 80 mesh and nylon 80 and 120 meshes can achieve high-efficiency filtration performance. Among them, nylon 80 mesh has the lowest filtration efficiency of 73%, and polyester 80 mesh has the highest filtration efficiency, which is 88.7%. When the production rate is 32 mm/s, no substrate can reach the standard of high efficiency filtration. Figure 9(c) shows the comparison of the air resistance of filter membranes produced at different production rates. As the production rate increases, the air resistance becomes smaller, mainly because the thickness of the nanofibers covered per unit area decreases.

In addition, the performance of the filter membrane of different substrates at different production rates is studied. We set the voltage to 60 kV, while the speed, the reciprocating speed of the nozzle, and the ambient temperature and humidity are unchanged. As shown in Figs. 9(a)–9(d), Fig. 9(a) is the comparison and analysis of the fiber diameter of different receiving substrates at different production rates. It can be found that the production rate has a certain effect on the fiber diameter. However, there is no obvious change law. According to Fig. 9(b), as the production rate accelerates, the filtration efficiency of the filter membrane becomes lower, which is mainly due to the decrease in the density of the nanofibers covered per unit area. Comparing the conditions of different receiving substrates, it can be found that the filtration efficiency of polyester 80 mesh is the highest at different production rates and reaches 98.23% filtration efficiency at 4 mm/s. Comparing the results of different production rates, it was found that under the condition of 4 mm/s, all filter membranes achieved high-efficiency filtration performance. Among them, at a production rate of 8 mm/s, polyester 60 mesh, polyester 80 mesh, nylon 80 mesh, and nylon 120 mesh have achieved high-efficiency filtration capabilities, while polyester 80 mesh has the highest filtration efficiency, reaching 95.5% filtration efficiency.

Select the best conditions for polyester 80 mesh, nylon 120 mesh, and nylon 80 mesh filter membranes at a production rate of 16 mm/s to prepare antibacterial test samples. Take polyester 80 mesh, nylon 120 mesh, and nylon 80 mesh filter membranes without nanofibers as the blank control group. Then, according to the test method of GB/T 20944.3-2008 Textiles—Evaluation for antibacterial activity—Part 3: Shake flask method, the antibacterial properties of the samples against Escherichia coli, Staphylococcus aureus, and Candida albicans are tested. As shown in Fig. 10, the report compares the antibacterial properties of three types of samples in different test bacteria. Among them, ATCC 29522, ATCC 6538, and ATCC 10231 correspond to Escherichia coli, Staphylococcus aureus, and Candida albicans, respectively. Therefore, the three types of samples showed excellent antibacterial properties in the three test bacteria.

An approach based on multi-needle electrospinning was applied for the first time to fabricate a PVDF electret air conditioning filter, achieving high filtration efficiency, low pressure drop, low air resistance, and antibacterial performance. In addition, we have been successful in the fabrication of the PVDF electret filter by multi-needle electrospinning equipment, using the PVDF solution modified with PVP to achieve a uniform fiber membrane without a beading structure. At the same time, the antibacterial performance of the PVDF fiber membrane can be realized by adding silver nitrate. It was confirmed that the fiber diameter becomes smaller when the spinning voltage increases because increased humidity is not conducive to the volatilization of the solvent and weakens the electric field strength, resulting in insufficient stretching and drying of the fiber. However, the diameter of PVDF fibers will decrease with the increase in the spinning voltage. The reason is that increasing the voltage will further strengthen the electric field strength and make the fiber more fully stretched and thinner. Moreover, after the fibers become thinner, higher filtration efficiency will be reached by the dense fiber layer. Besides, the filtration efficiency test and fiber distribution proved that the uniformity of fiber deposition will be better with the increase in the polyester mesh number and the higher filtration efficiency tested by the filter. By multi-needle electrospinning, we can maintain polyester 80 mesh, nylon 120 mesh, and nylon 80 mesh filter membranes with high filtration efficiencies at a production rate of 16 mm/s.

All authors contributed equally to this work.

The data that support the findings of this study are available from the corresponding author upon reasonable request.

This work was financially supported by the China Postdoctoral Science Foundation (Grant No. 2018M633010), the University Innovation and Entrepreneurship Education Major Project of Guangzhou City (Grant No. 201709P05), the Guangdong University Students Science and Technology Innovation Cultivation Special Fund (Grant No. pdjhb0160), and the Key Laboratory Construction Projects in Guangdong (Grant No. 2017B030314178).

The filtration efficiency of filters was evaluated using a filter tester (Model 8130, TSI, USA). The filter material testing instrument uses the principle of counting method. Aerosol particles are generated by a particle generator, and the compressed gas after drying and purification is sent to the test port. The upstream dust particle counter measures the number of particles at the test port. Under the conveyance of compressed gas, most of the aerosol particles are intercepted and adsorbed by the filter material after passing through the filter material, and a small part is discharged downstream, detected by a sub-counter, and then calculated. Using Eq. (A1), we can thus obtain the corresponding value,

where P_i is the particle group counting efficiency, N_1i is the average value of the technical concentration of the particles with an air inlet greater than or equal to a certain size (particles/l), and N_2i is the average value of the technical concentration of particles with an air outlet greater than or equal to a certain particle size (particles/l).

Air resistance is another core parameter of filters, also known as the permeability of filters, and is an important indicator for measuring the gas circulation performance of the filter material. Meanwhile, air resistance and filtration efficiency are related to each other. Generally, the higher the filtration efficiency, the greater the resistance. However, with the development of high-efficiency and low air resistance filter membranes, the production of fluffy and porous three-dimensional nanofibers and electret nanofiber membranes is complete. According to Eq. (A2), the air resistance of the filter material can be determined by the pressure drop ΔP of the filter material,

where P₁ is the gas pressure at the filter inlet and P₂ is the gas pressure at the filter outlet.

For unused filter materials, the ΔP value is determined by the gas flow rate and the porous material characteristics of the filter material. When the filter material is used for filtration, the value of ΔP is also related to the particulate matter intercepted by the filter material. Generally, during the use of filters, the pressure drop ΔP increases with the increase in the particulate matter deposited on the filter material.

The automatic filtration efficiency tester is used to test the air filtration performance of the fiber membrane. The test results include filtration efficiency (η) and filtration pressure drop (P). The quality factor (QF) defined by the comprehensive performance index is used to evaluate the filtration performance of the nanofiber membrane. The larger the QF value, the better the filtration performance of the fiber membrane,