Air Filtration with Electrospun Nanofibers

There is a wide range of application which an air filter media can be used from industrial to personal and household. Industrial usage of air filter media includes clean room and fossil-fuel power station. Ships and vehicles also require air filter media to reduce particulate emission from their exhaust. Personal and household usage of air filter media includes facemask, air purifier, vacuum cleaners and window filter screen. Air filter is the first commercial application of electrospun product with the industrial production of Petryanov filters in USSR in 1939. It is only until the early 1980s that a company from US (Donaldson) begins making and selling filter media with electrospun nanofiber fibers. Little is known about the enhanced performance of the air filtration media with the addition of nanofibers in the public domain until the early 2000s when researchers start to carry out experiments in this area of application. In a demonstration of the filtration effectiveness of electrospun nanofibers membrane, a thin coat of electrospun nanofibers over a metal mesh has been shown to collect more than 95% percent of PM_2.5(Particulate matter, 2.5 µm) in a polluted city environment while allowing 90% percent transparency [Liu et al 2015].

An electrospinning setup of by Bioinicia [Video of FLUIDNATEK LE-50].

To understand how electrospun fibers significantly enhance the filtering performance of the filter media, it is important to understand the particle capturing mechanism. Depending on the size of the particles, there are four main mechanisms which the filter media removes them. Particle size larger than the pore size of the filter media are simply captured by sieving mechanism. For smaller particles, the remaining three mechanisms are inertial impaction, interception and diffusion. Inertial impaction occurs when the inertia of the particle as it is carried by the air stream hits the fiber although the air stream has flowed round the fiber. When the particle is sufficiently small to flow with the air stream without leaving its path, it may still come into contact with the fiber surface as it attempts to flow round it and be captured (intercepted) by the fiber. For very small particles (< 100 nm), its movement is dominated by Brownian motion and it is captured by the fibers through random collision.

The efficiency of nanofibers in capturing nanoparticles from various smoke sources by collision was shown in a study by Kusumaatmaja et al (2016). Using electrospun polyvinyl alcohol nanofiber membrane, smoke from cigarette, waste combustion and vehicle combustion was passed through it. Observation under SEM showed clearly, nanoparticles adhering to the surface of the nanofibers. Removal of the smoke particles is not through size exclusion as the particles captured are much smaller than the pore size.

SEM images of electrospun PVA membrane (a) before filtration; after filtration of (b). cigarette smoke, (c) smoke from waste combustion, (d) smoke from vehicle combustion [Kusumaatmaja et al 2016].

The Covid-19 epidemic in 2020 has brought much attention to the importance of facemasks for both health care workers and the general public. In particular, the sudden surge in demand for facemasks have and shortage has brought to attention other considerations in facemask design such as reusability. As of 2020, almost all N95 certified facemasks are made using melt-blown filtrate. Greater research on electrospun nanofibers for use as facemasks has shown some interesting results. Ullah et al (2020) did a comparison study on the reusability of filters made from melt-blown (MB) polypropylene (PP) and polyvinylidene difluoride (PVDF) bonded on spun bond PET (Polyethylene terephthalate). Filtration efficiency of MB filter dropped significantly to around 65% or less after 10 cycles of ethanol spraying or 24 h ethanol dipping. With the nanofiber filter, the filtration efficiency is maintained at 98% regardless of ethanol cleaning method. The difference in the filtration efficiency after cleaning may be attributed to the filtering mechanism. For MB filters, static charge plays an important role in its filtering capability. Therefore, repeated washing would substantially reduce its static charges. For nanofiber membranes, its mail filtration mechanism comes from particles impacting the fiber and size exclusion. Therefore, repeated washing does not adversely impact its performance. Other advantages of nanofiber filter membrane are its quick drying time of 10 min compared to 3 h for MB filters and greater user comfort as it is thinner which reduces heat retention within the mask. In terms of reusability, nanofiber filter membrane has demonstrated several significant advantages over conventional single use MB filters.

Description	Illustration
Sieving / Size exclusion Mechanism Particle size: Larger than pore size of membrane Particle flight path: Influenced by air stream but does not closely follow its path due to inertia
Inertial Impaction Mechanism Particle size: Micrometer range Particle flight path: Influenced by air stream but does not closely follow its path due to inertia
Interception Mechanism Particle size: Low micrometer to sub-micrometer range Particle flight path: Follows air stream
Diffusion Mechanism Particle size: Nanometer range Particle flight path: Brownian motion

Fiber diameter on filtration efficiency

An important difference between nanofiber and larger diameter fiber filter media is the phenomena of slip flow. For larger diameter fiber, the air molecules colliding on the surface of the fiber will acquire the velocity of the fiber (stationary) plus its random motion due to their thermal energy. This results in a pressure drop due to the presence of the fibers. For nanofibers, less air molecules actually impact the fiber surface thus they are able to better maintain their initial flow velocity which is known as slip flow. This slip flow results in less pressure drop compared to larger diameter fibers. It has been shown that pressure drop across nanofiber membrane in relation to its solid volume fraction is less than model prediction without slip flow [Hosseini et al 2010]. Studies have also shown significantly less pressure drop for nanofiber filter media compared to microfiber filter media of the same filtration efficiency [Park et al 2005]. With slip flow, more particles will travel near to the fiber resulting in higher diffusion, interception and inertial impaction. Significant improvement in dust reduction has been recorded when a standard cellulose filter media was coated with a nanofiber layer [Graham et al 2002]. While filtration efficiency increases with greater solid volume fraction (of fibers), this also leads to a greater pressure drop despite slip flow effect of nanofibers. Zhao et al (2016) showed that electrospun polyacrylonitrile (PAN) diameters between 60-100 nm was the most effective in facilitating the slip flow effect which coincides with the mean free path of air molecules at 65.3 nm. To determine the optimum balance between these factors, a Figure of Merit (or Quality Factor) is used.

FOM = -In(P)/Δp

Where P is the penetration of particle; Δp is the pressure drop.

The penetration of particle, P = (particle concentration downstream of the filter) / (particle concentration upstream of the filter).

Using Figure of Merit to determine the performance of the filter media, several studies have shown that smaller fiber diameter (< 1 micron) is better for filtering particles of diameter more than 100 nm. For particles of smaller diameter, the performance from larger diameter fiber is better [Wang et al 2008, Hosseini et al 2010]. A study by Matulevicius et al (2014) using electrospun polyamide 6 (PA 6) and polyamide 6/6 (PA 6/6) nanofibers to demonstrate the effect of fiber diameter on the filtration performance on 100 nm and 300 nm diameter polystyrene latex particles. For 300 nm particles, filtration efficiency was highest (91%) on electrospun PA 6/6 nanofibers which has diameter of about 65 nm. For electrospun PA 6 fibers with diameter of about 230 nm, filtration efficiency was similar for both 100 nm and 300 nm diameter particles at close to 87%. It is possible that with smaller fiber diameter, the pore size is smaller and this improves its ability to trap larger size particles. When the particle size is smaller, size exclusion by the membrane pore size is no longer effective and thus there is less difference in the filtration efficiency between fiber diameter less than 100 nm and fiber diameter less than 300 nm.

When the average fiber diameter increases to 300 nm, the filtering efficiency of the membrane drops significantly [Li et al 2006]. Model simulation using a 3-D structure resembling nanofiber (diameter less than 200 nm) filter media, low air flow and no-slip boundary conditions, nanoparticles less than 100 nm are shown to be mostly captured through Brownian diffusion while larger particles are captured through interception with the fiber. Cake formation for nanoparticles is found mainly in the middle of the nanofibrous membrane since Brownian diffusion is the dominant capture mechanism [Maze et al 2007]. Experimental results using fibers from meltblowing showed that with smaller fiber diameter, the quality factor increases for all particles from 10 to 100 nm. It is only for particles smaller than 30 nm where the quality factor for 1.3 µm diameter fibers membrane is lower than the base layer [Podgorski et al 2006]. In the study of tar removal efficiency by Molaepour et al (2014) using electrospun fiber membrane, filtration efficiency by fibers with diameter of 280 nm is significantly (10% and greater) better than fibers with diameters of 370 nm and 620 nm. Another model using 300 nm diameter fibers also showed that factors that lead to reduced Brownian motion intensity (increased air velocity, viscosity and air pressure) results in a reduction in filtration efficiency for particles size less than 200 nm while increase in air temperature and particle-fiber friction coefficient increases filtration efficiency through increased Brownian motion intensity and reduced particle slip respectively [Sambaer et al 2011]. A study by Ahn et al (2006) using 200 nm diameter fiber showed that its filtration efficiency drop marginally from 99.97% to 99.96% for 300 nm particle when the face velocity increases from 3 m/min to 6 m/min. Electrospun membranes with fiber diameter less than 300 nm may not always perform better than membranes with larger diameter fibers. Beckman et al (2023) constructed an air filter membrane using electrospun polyacrylonitrile (PAN) membrane stabilized for 240 min at 270 °C. Membranes with fiber diameter of 216 nm and 462 nm were prepared. With airflow velocity of about 5 cm/s, the FOM of the PAN membrane with fiber diameter of 216 nm and 462 nm was 0.0137 and 0.0250 respectively. This result is contrary to the predicted better performance of membranes with smaller diameter fibers. Initial filtration efficiency of 216 nm fiber diameter membrane and 462 nm fiber diameter membrane was 99.8% and 97,8% respectively but the pressure drop was 459 Pa and 154 Pa respectively. The large increase in pressure drop of the smaller diameter fibers contributed to its lower FOM value. Stabilization of the PAN membrane may have led to compaction of the fibers with the effect being more significant for smaller diameter fibers.

Filter particle size

Most air filter media is made out of nonwoven microfibers. Due to their larger pore size, initial filtration efficiency is low and is ineffective for particles size in the range of 100 to 800 nm. With nylon nanofiber (diameter 120 nm) coating of 0.1 g/m², the filtration efficiency of particles less than 1 microns rise to more than 80% as compared to less than 50% for non-nanofiber coated filter media [Li et al 2006].

Fiber diameter	Weight / Thickness of fiber	Filtration Efficiency	Pressure drop	Reference
300 nm diameter fiber	16.48 g/m2	98% for 0.3 µm particles	13.27 mmH2O at 3 m/min of face velocity	Park et al 2005
Glass fiber filter media	81.46 g/m2	98% for 0.3 µm particles	37.05 mmH2O at 3 m/min of face velocity	Park et al 2005
HEPA filter	500 µm, 78.2 g/m2	99.97% for 0.3 µm particles	< 25 mmAq at 3 m/min	Ahn et al 2006
Nylon 200 nm diameter fiber	50 µm, 5.75 g/m2	99.97% for 0.3 µm particles	> 25 mmAq at 3 m/min	Ahn et al 2006
Nylon 200 nm diameter fiber	100 µm, 10.75 g/m2	99.993% for 0.3 µm particles	80 mmAq at 3 m/min	Ahn et al 2006
91 nm diameter fiber on filter media	2.3 µm thickness nanofiber	92% for NaCl aerosols with median diameter of 75 nm	40 mmH2O at 32 L/min	Li et al 2013
120 nm diameter fiber on filter media	0.1 g/m2	~88% for 1 µm particles	7.87 mmH2O at 0.3 m/min	Li et al 2006
120 nm diameter fiber on filter media	0.1 g/m2	~60% for 0.3 µm particles	7.87 mmH2O at 0.3 m/min	Li et al 2006
300 nm diameter fiber on filter media	0.1 g/m2	~35% for 1 µm particles	7.87 mmH2O at 0.3 m/min	Li et al 2006
Original filter media	NA	50% for 1 µm particles	NA	Li et al 2006
Original filter media	NA	~11% for 1 µm particles	NA	Li et al 2006
120 nm diameter fiber on filter media	0.1 g/m2	~74% for 0.3 µm particles	~50 Pa at 6 m/min	Heikkila et al 2008
120 nm diameter fiber on filter media	0.1 g/m2	>90% for 1 µm particles	~50 Pa at 6 m/min	Heikkila et al 2008

With air pollution, the emphasis has been on PM2.5 filtration efficiency. However, with Covid-19 pandemic in 2000, there has been a shift in attention towards filtering virus particles which are much smaller. Blosi et al (2021) used electrospinning to construct filter membranes made of polyvinyl alcohol (PVA) load with silver nanoparticles for antibacterial property. For filtering out virus size particles, a sample with 5 layers of the membrane was prepared. NaCl aerosol particles comparable to a virus size distribution with median diameter around 70 nm and mass median diameter of 500 nm were tested on the filter sample. A high filtration efficiency of more than 95% was recorded for all particle sizes at filtration velocity of 5.5 cm/s and this was reduced to a minimum of 80% for filtration velocity of 16.7 cm/s. Pressure drop and filtration efficiency values are within range of FFP1 and FFP2 masks, even at the highest filtration velocity of 16.7 cm/s which makes the five layered sample suitable for use as facemasks.

Pore Size

A less investigated parameter on slip flow is the pore size of the membrane. The pore size of electrospun membrane is formed by overlapping fibers. Many studies have shown that a reduction in fiber diameter would cause a corresponding reduction in pore size. While reduction of fiber diameter has the positive effect of facilitating slip flow and thus reduce pressure drop, Zhao et al (2016) has shown that when electrospun polyacrylonitrile (PAN) nanofibers diameter is below 60 nm, the pressure drop increased significantly. This has been attributed to air flow distribution around the periphery of the fibers where interference of air flow from neighbouring fibers become significant.

Fiber Morphology

Electrospinning is known to produce fibers with different structures depending on the parameters and solution used. Most investigation of air filtration on electrospun membranes is based on smooth fibers. However, with a different fiber morphology, the air flow will be different and the corresponding filtration characteristics can be expected to be different as well. Zhou et al (2022) showed that with the addition of sodium dodecyl sulfate (SDS) to polyvinylidene fluoride (PVDF) solution, fibers with smaller diameters start to branch off thicker fibers. As SDS increases the conductivity and reduces the surface tension of the solution. Greater electrification of the electrospinning jet and lower surface tension encourages side branches to erupt from the main electrospinning jet. Filtration performance was carried out using NaCl particles with a diameter of 0.3 µm. The filtration performance of SDS/PVDF electrospun membrane was consistently higher than pure PVDF electrospun membrane of the same PVDF concentration. The pressure drop of SDS/PVDF electrospun membrane was consistently lower than pure PVDF electrospun membrane of the same PVDF concentration leading to a higher quality factor of SDS/PVDF electrospun membrane. The presence of both coarse and fiber fibers uniformly distributed in the membrane may have improved air permeability while the fiber fibers were able to capture fiber particles thus resulting in a lower pressure drop and higher filtration efficiency.

SEM images of PVDF/SDS nanofiber membranes with different SDS concentrations: (a) 0% wt, (b) 0.1 wt%, (c) 0.2 wt%, and (d) 0.3 wt% [Zhou et al 2022].

In most cases, having beads on the electrospun fibers are thought to be undesirable. However, in a study by Akmal et al (2019), they found that having beads of fibers may actually improve the air filter quality factor. Using electrospun acrylonitrile butadiene styrene (ABS) fiber membrane, and tests with incense smoke, they found that the membrane with the highest air filter quality factor had the highest beads density. They hypothesized that the beads create a physical separation between fiber layers and this allows easier air flow through the membrane hence reducing the pressure drop without substantially reducing air filtration efficiency. Beaded fibers have a much smaller fiber diameter than smooth fibers and this may have compensated for the otherwise drop in efficiency. However, it must be noted that the presence of beads reduces the mechanical strength of the membrane. They also found that membranes with beaded fibers clogged faster than smooth fiber membrane.

Membranes with beaded fibers generally exhibit greater hydrophobicity and this may also benefit air filtration performance. Liu et al (2022) constructed a superhydrophobic membrane with electrospun polyvinyl alcohol(PVA)/Eugenol (Eo) fibers and electrosprayed ethyl cellulose (EC)/Eugenol (Eo) beads. Eo is a hydrophobic aromatic compound which may increase the hydrophobicity of electrospun PVA fibers and electrosprayed EC beads. PVA is a green, non-toxic, and degradable linear polymer, however its hydrophilic property makes it unsuitable for use in high humidity environments. By electrospraying a layer of EC/Eo beads on the base layer of electrospun PVA/Eo fibers, the water contact angle of the PVA membrane increases from 142.8 to 151.1°. The composite membrane at optimum electrospinning and electrospraying duration showed a low filter pressure drop of 168.1 Pa with high filtration efficiencies of 99.74 and 99.77% for PM1.0 and PM2.5 respectively. The respective quality factors were found to be 0.0351 and 0.0358 Pa^-1. In a high relative humidity condition of 90%, there is a slight decrease in filtration efficiency from 99.95 at 15% to 99.67%.

Other forms

Electrospun nanofibers for used in air filtration need not be restricted to the form of 2D membrane. Sponge-like filter media may be used as depth filters for greater particulate holding capacity. This is particularly useful in building filtration system where the filter media are replaced every few months. There are several methods for fabricating 3D block fibrous structure. Deuber et al (2016) used a combination of short nanofibers suspension and controlled freeze drying to construct a sponge comprising of macropores from the freeze drying and micropores from the distance between nanofibers. Their preliminary study showed that the filtration efficiency increased from 91% to 99.96% when the macropores were reduced from 122.6 µm to 15.2 µm. However, the most penetrating particle size (MPPS) remained from 134 to 202 nm across the sponges with different macropore sizes.

Another way of increasing filtration efficiency is to give the filter a surface charge. Zhang et al (2017) investigated the surface charge characteristic of polyacrylonitrile (PAN) and PAN with TiO₂ and its effect on particle removal. With the presence of TiO₂ in electrospun PAN fibers, the charge retention of the filter were significantly better than pure electrospun PAN filter membrane. Surface voltage of electrospun PAN fibers with 2% TiO₂ showed a voltage decrease by 38% after 96 hrs but pure electrospun fibers had a voltage decrease of 57%. With increasing amount of TiO₂, the voltage decrease were reduced. Zhang et al (2017) hypothesized that the reduction in fiber diameter with increasing TiO₂ content may have helped in the charge retention. KCl particles of 0.3-0.5 µm penetration was also reduced with greater surface voltage.

[+]

[-] comments powered by Disqus

Content [Air Filtration]