Performance of Electrospun Polyvinylidene Fluoride Nanofibrous Membrane in Air Filtration

2.2.1 Solution preparation

The PVDF powder was dissolved in a solvent of DMF at room temperature to obtain homogeneous solutions with concentrations of 6%, 8%, 10%, 12%, and 14%.

2.2.2 Electrospinning conditions

The PVDF nanofibrous membrane was prepared by needle electrospinning equipment (provided by Beijing Ion Beam Technology Co., Ltd., mode: WL-2C). The electrospinning equipment consisted of two plastic syringes and two steel needles with an inner diameter of 0.4 mm. The needles were connected to a high voltage power supply, which can be regulated to produce uniform voltages. Typically, electrospinning was performed at a voltage of 13 kV, at a distance of 15 cm between the needle tip and the grounded collector, at a solution feed rate of 0.5 ml/h, at a temperature of 22 ± 3°C, and at a humidity of 50 ± 5%. The electrospinning process lasted for 10 hours, and the resultant electrospun nanofibers were collected on a glass fiber grid.

2.2.3 Testing methods

The viscosity of solutions was tested by a rotary viscometer at a temperature of 20 ± 2°C and a relative humidity of 60 ± 5%. The electrical conductivity of solutions was measured by a conductivity meter (Shanghai Optical Instrument Factory, mode: DDSJ-308A) at a temperature of 20 ± 2°C and a relative humidity of 60 ± 5%. The mass of nanofibrous membrane was measured by ME204E electronic balance, and then, the areal weight was calculated by getting the area of the nanofibrous membrane. The fiber diameter and its morphology were observed by a thermal field emission scanning electron microscope, and the diameter of fibers was analyzed using Image-Pro Plus software.

The pore size of nanofibrous membrane and its distribution were analyzed by the CFP-1500AE type aperture analyzer, in which the bubble point method was applied. The principle is that the sample is completely infiltrated with a wetting agent of known surface tension and then placed in the sample chamber. Under a certain pressure, the gas passes through the capillary pores when the sample is in a dry state and in a wet state. During the whole testing process, the pressure and air flow change, and the pore size and distribution of the sample were finally calculated.

The porosity of nanofibrous membrane was obtained from the following equation:

where ɛ is porosity and ρ_m and ρ_f are densities of nanofibrous membrane and fiber, respectively. ρ_f was taken as 1.70 g/cm³.

The air permeability of nanofibrous membrane was evaluated using YG461E-III automatic air permeability meter. The test pressure was 50 Pa, and the area of testing head was 20 cm². The average values and standard deviations were determined from five repeated measurements for each sample.

The filtration efficiency of nanofibrous membrane was evaluated using TSI8130 automatic filter tester, which is able to generate both paraffin oil and NaCl aerosol as testing media. In addition, the filter tester features a high degree of automation and self-diagnostics that greatly improve overall measurement performance. The technical performance of the tester was according to the technical requirements of GB2626-2006 “Respiratory protective equipment self-priming filter type particulate respirator”.

3.2.1 Fiber diameter and its morphology

The surface morphology of the fibers was observed using a thermal field emission scanning electron microscope. The diameters of the fibers at different solution concentrations were also analyzed using Image-Pro Plus software. The results are shown in Figure 3.

Figure 3

Fiber morphology and its diameter distribution. (A) 6% solution concentration, (B) 8% solution concentration, (C) 10% solution concentration, (D) 12% solution concentration, and (E) 14% solution concentration.

The fibers were able to be electrospun from the given solution concentrations. However, the beaded fibers appeared when the solution concentrations were 6% and 8%, which made the fiber very uneven in diameter (Figure 3a and b). With the increase in solution concentration, the beads on fibers disappeared, which led to relatively uniform single fiber in diameter (Figure 3d and e). Besides, the surface of nanofibers was very smooth when the solution concentration was less than 12% (Figure 3a–d), while the surface of fibers from 14% solution concentration was much rougher and had bumps on it (Figure 3e). These rough bumps might be due to gold-plated particles that adhered on fibers when gold was plated before testing. Additionally, the distribution of fiber diameter is given in Figure 3 as well. The fiber diameter ranges were from 20∼ to 70 nm, 20∼ to 80 nm, 20∼ to 120 nm, 100 to ∼1,790 nm, and 240 to ∼2,450 nm for fibers from different solution concentrations.

The average diameters of PVDF nanofibers at different solution concentrations are given in Figure 4. The increase in average fiber diameter was along with the increase in solution concentration. However, there was a significant increase in fiber diameter when the solution concentration changed from 10% to 12%, responding to a change in fiber diameter from 51.39 nm to 621.05 nm. Meanwhile, the fiber diameter distributions were at quite narrow range when the solution concentration was less than or equal to 10%.

Figure 4

Average diameter of nanofibers at different solution concentrations.

Apart from the influence of solution concentration on fiber diameter, Angammana and Jayaram stated that the increase in electrical conductivity of the solution resulted in a significant reduction in fiber diameter because the jet carries more charges [21]. Therefore, the experimental results here indicated that the solution concentration played a more dominant role than electrical conductivity in influencing fiber diameter.

3.2.2 Porosity of PVDF nanofibrous membranes

The areal weight and thicknesses of PVDF nanofibrous membranes were measured, and then, the porosity of nanofibrous membranes was obtained according to equation (1). The results are given in Table 1.

3.2.3 Pore size distribution of nanofibrous membrane

The pore size of nanofibrous membrane from 6% solution concentration was not able to be evaluated from our current testing equipment since the pore size was too large. The pore size of nanofibrous membrane from 8% solution concentration was mainly distributed from 20 μm to 380 μm (the interval was 10 μm), and the majority of pore size was from 20 μm to 50 μm and from 160 μm to 380 μm, and the percentage of these pore size interval was very close (Figure 5A). When the solution concentration reached to 10%, the pore size of nanofibrous membrane decreased dramatically (Figure 5B–D). The pore sizes of nanofibrous membrane from 10% to 14% solution concentrations were from 1.2 μm to 1.5 μm, 2.4 μm to 4.5 μm, and 2.4 μm to 5.0 μm.

Figure 5

Pore size distributions of nanofibrous membranes. (A) 8% solution concentration, (B) 10% solution concentration, (C) 12% solution concentration, and (D) 14% solution concentration.

The differences in pore size could be explained by combining fiber diameter, thickness, and porosity of nanofibrous membrane. The diameters of fiber from 6% and 8% solution concentrations were much smaller than those from 12% and 14% solution concentrations, and their porosities were relatively higher (Figure 4 and Table 1). Even the diameter of fiber from 10% was small; however, its porosity and thickness were also small, which meant a more compact structure and smaller pore size.

The mean pore sizes of nanofibrous membranes were decreased first and then increased, and the values were 237.03 μm, 1.27 μm, 3.24 μm, and 3.24 μm, respectively, when solution concentrations were from 8% to 14% (shown in Figure 6), which can also be observed from Figure 5. The results indicated that the nanofibrous membrane produced from 10% solution concentration had most compact structure, which could be due to the lower porosity (compared with 6% and 8%) and the smaller fiber diameter (compared with 12% and 14%).

Figure 6

Mean pore size of nanofibrous membranes at different solution concentrations.

3.3.1 Air permeability

The air permeability can indirectly reflect the airflow resistance of nanofibrous membrane. The larger the air permeability, the lower the airflow resistance. Theoretically, the air permeability of fibrous filter is a function of pressure drop, filter thickness, fiber diameter, and porosity [22,23,24]. However, the experimental results in this work did not match the predicted values to a great extent, which might be due to the nanoscale fiber diameter and very small pore size. The air permeability of nanofibrous membrane is given in Table 2.

3.3.2 Filtration efficiency

The filtration efficiency of fibrous filter can be calculated using the following equation [25], which is based on the concept of single fiber filtration efficiency:

where E is the filtration efficiency of fibrous filter, η is the packing density of the fibrous filter, η is the single fiber efficiency, L is the filter thickness (m), and d_f is the fiber diameter (m).

Considering the particle capture mechanisms, the single fiber efficiency, η, can be expressed as

where η_I is the fiber efficiency by inertial impaction, η_R is the single fiber efficiency due to interception, η_D is the single fiber efficiency due to convective Brownian diffusion, and η_DR is the single fiber efficiency due to interception of diffusing particles.

η_I, η_R, η_D, and η_DR can be calculated as [25,26,27]

where S_tk is the Stokes number, Ku is the Kuwabara hydrodynamic factor, a is the packing density, R is the interception parameter, Pe is the Peclet number, and Kn_f is the fiber Knudsen number. These parameters are calculated as [25,26,27].

where ρ_p is the particle density (in this experiment, NaCl aerosol particle was used and its density is 1.333 g/cm³), d_p is the particle diameter (0.3 mm), U_p is the particle velocity relative to the gas flow (m/s), C is the slip correction factor, μ is the viscosity of the gas (17.9 ′ 10⁻⁶ Pa×s), B is the mechanical mobility of the particle, m is the particle mass (kg), and λ is the mean free path of the gas molecules (the mean free path is about 0.0645 mm for air molecules at 293 K and normal atmospheric pressure).

where d_f is the fiber diameter (nm), D is the particle diffusion coefficient, U is the flow velocity (m/s), k is the Boltzmann's constant (1.3805 ′ ′10⁻²³ J/K), and T is temperature (K).

Comparing with the filtration efficiency from experiments, the predicted values from the mathematical model were higher, especially when the fiber diameter was 36.8 nm (Figure 7). Therefore, the ratio of particle diameter to fiber diameter was more than 8, which was beyond the range for the theoretical model.

Figure 7

Comparison of filtration efficiency of fibrous membrane from experiment and from model.

3.3.3 Quality factor

Quality factor (QF) is usually used to evaluate the comprehensive performance of filters, which can be obtained from the following equation:

where E is the filtration efficiency and ΔP is the pressure drop (Pa).

The QF was closely related to pressure drop, and they had a negative correlation with each other. The QFs of samples were decreased with the increase in solution concentration (Figure 8). The QF of fibrous membrane produced from 6% solution concentration was higher than all others, and the high QF was contributed by the very low pressure drop (about 5 Pa). However, this sample cannot be taken as the best one since its filtration efficiency (less than 50%) was too low to meet the application requirements. Instead of fibrous membrane from 6% solution concentration, the membrane from 10% solution concentration had a very high filtration efficiency and relatively low pressure drop. Together with the fiber morphology, this sample could be considered as the optimum one.

Figure 8

Filtration efficiency, pressure drop and quality factor of samples.