key: cord-1049643-w50c5kjn authors: Zhou, Mengjuan; Shi, Lulu; Dai, Hongyu; Obed, Akampumuza; Liu, Penghong; Wu, Jiajun; Qin, Xiaohong; Wang, Rongwu title: Facile fabrication of reinforced sub-micron fibrous media with hierarchical structure compounded thermally for effective air purification in application date: 2022-02-24 journal: Sep Purif Technol DOI: 10.1016/j.seppur.2022.120726 sha: 9d5b5da33270047c1e4bb03e2aa06cccb9cd34ae doc_id: 1049643 cord_uid: w50c5kjn Air pollution has steadily worsened in recent years, and the coronavirus disease 2019 has been spreading since 2020. The electrospun fibrous filters present superior filtration performance, while the low mechanical property and yield of them limit their application, which must be addressed urgently. Herein, polyacrylonitrile (PAN) sub-micron fibrous membrane with hierarchical structure was easily manufactured using free surface electrospinning in mass production for air purification. The “sandwich” structured fibrous filter was thermally bonded with bi-component nonwoven through traditional bonding procedures, due to melting and bonding of the cortex of bi-component fibers, in which the electrospun fibrous web as the mid layer with tortuous channels showed superior filtration performance for aerosol particles with diameter of 260 nm, which could effectively intercept different-sized particles suspended in the air. In addition, the impact of the processing parameters on the characteristics and filtration mechanisms of thermally bonded composite materials was thoroughly investigated. The results showed that composite material with “dendrites” and “axon” morphologies presented the best formability, outstanding peeling strength and breaking strength, and steady filtration performance, following an easy through-air bonding procedure, making it useful for post-processing in air purification. The reinforced composite filter, which is thermally bonded with sub-micron fibers with high yield and nonwoven, is save-energy and has a low operation cost, indicating promising commercial possibilities. Air pollution is one of the most significant sources of environmental pollution, and it has grown 24 in prominence as a result of industrialization, attracting increased societal attention in recent 25 years. [1] Taking an example of suspended fine particulate matters (PMs) with an aerodynamic 26 diameter of less than 2.5 μm (PM 2.5 ), these can easily penetrate the human lungs, increasing the 27 incidence of respiratory and heart diseases. [2, 3] In addition, there is an increasing risk posed 28 by the emergence of respiratory infections, for instance, the sudden emergency and rapid spread 29 of the coronavirus disease 2019 (COVID- 19) , whose primary dissemination route is droplet 30 transmission.[4-6] As a result, the professional anti-haze masks are required to reduce or 31 eliminate the risk posed to the human body by air pollutants (aerosols and viruses). Furthermore, 32 the performance of precision instruments, as well as the atmosphere of enclosed public spaces 33 such as hospitals and airports, are heavily reliant on the efficient air purification. Thus, the 34 deployment of air systems with superior filtration properties would ensure the safety of human 35 and the unintended performance of such equipment. [7, 8] 36 Electrospinning process produces a disordered accumulation of sub-micron or micron fibers, 37 resulting in a highly porous membrane with distinct properties. [9, 10] This is characterized by 38 a uniform packing density and tunable network geometry, both of which are conducive to 39 filtering fine particles.[11-13] The arrangement and multi-level structure of fibers with different 40 fiber diameters show a positive effect on the filtration performance of fibrous filters. [14, 15] 41 While, only scanty work is available on the fabrication of sub-micron fibrous webs with 3 1 hierarchical structure as an air filter. In practice, electrospun sub-micron fibrous materials with 2 low strength need to be combined with non-woven fabric to meet mechanical strength 3 requirements. However, due to the low interfacial force and bonding strength between the sub-4 micron fibrous membranes and the substrate layer, they may easily peel off or fray during post-5 processing (composite processing and filter preparation) and in the actual applications. 6 However, if the two layers were superimposed directly, the product's safety would be 7 jeopardized. The main effective ways to improve the adhesion force between them are chemical 8 and thermal bonding processes. The thermal bonding process takes advantage of the 9 thermoplastic properties of polymers, in which the polymer softens, diffuses, flows, adheres 10 between the fibers, and finally solidifies after cooling to form a bonding point. This method is 11 widely used in the preparation and processing of non-woven materials because the finished 12 products, which do not contain chemical adhesives, are environmental-friendly and effective. 13 The thermal bonding process can be divided into three categories, including extra-sonic 14 composite method [16] , through-air method [17] , and hot-rolled bonding method. [18] Among of 15 them, the through-air and hot-rolled bonding processes are used commercially. In the through- 16 air bonding process, the nonwoven fabrics with a thermoplastic property are heated with hot air 17 to reach the melting point, melted to produce viscous flow, and then condense at the fiber 18 intersections. As for the bonding process, fibrous materials are heated and pressurized by one 19 or two pairs of heated steel rollers, causing some fibers to melt, flow, diffuse, bond, cool down, 20 and be strengthened, which does not require a binder preparation station and is energy-efficient, 21 low-cost, and widely used in actual production. [18, 19] Previous research on electrospun 22 fibrous materials as air filters focuses primarily on the front-end electrospinning preparation 23 and materials' properties, ignoring the subsequent composite process, which is critical in 24 practical applications. Nguyen and Kim [20] chose a commercial heat roller to manufacture 25 PAN nanofiber air filter laminated with non-woven fabrics using heat and pressure, which 26 showed the material with a sandwich structure (non-woven fabric, PAN nanofiber, non-woven 27 fabric) possessed a high filtration efficiency with structural stability, while the work mainly 28 focused on optimizing the electrospinning conditions. Till now, there is a rare detailed report 29 about the influence of the bonding composite process on the properties and air-filtration 30 performance of the electrospun fibrous filter. 31 In this study, a PAN sub-micron fibrous membrane with hierarchical structure was simply 32 prepared via free surface electrospinning with large-scale production as air filter cartridge ( Fig. 33 1(a)), in which the upper and lower substrates were polyethylene/polypropylene (PE/PP) bi-34 component composite hot-air nonwoven fabric, preparing a "sandwich"-structure compound 35 fibrous filter (nonwoven fabric/sub-micron fibers/nonwoven fabric). First, the effect of fiber 36 diameter, areal density, and fibrous web structure on the filtration performance of PAN fibrous 37 membrane was investigated. Then, the composite materials resembling a sandwich structure 38 were compounded via two different thermally bonding processes (i.e., hot-rolled and through-39 air bonding processes) ( Fig. 1(b, c) ), respectively, to improve interlaminar fracture toughness 40 and suppress the delamination in laminated non-woven fibers. Furthermore, the impact of 41 process parameters (calender temperature and feeding speed, as well as nip pressure in hot-42 rolled process; heating temperature and time in through-air bonding process) on the properties 43 of composite materials was investigated in order to ensure excellent filtration performance and 44 mechanical properties for post-processing in air purification, which was conducive to practical 4 1 application. Polyacrylonitrile (PAN, M w = 85,000) with a material density of 1.184 g cm -3 was purchased 5 from Shanghai Chemical Fibers Institute. N-dimethylformamide (DMF, AR) was supplied by 6 Sinopharm Chemical Reagent Co., Ltd., China. Hot-air non-woven fabric of sheath-core 7 polyethylene/polypropylene (PE/PP) bi-component composite fibers (the areal density is 25 g 8 m -2 , the sheath/core ratio is 1: 1, and the fiber diameter is 18 μm) possesses the negligible 9 filtration efficiency (5%) and pressure drop (4 Pa) at airflow velocity of 5.3 cm s -1 for collecting 10 the electrospun sub-micron fibers, provided by Sanming Kangerjia sanitary products Co., Ltd., 11 Fujian. All reagents were used without further purification. 12 13 PAN powder was dried in an oven for 3 hours, then it was dissolved in DMF and stirred 14 magnetically at room temperature for 24 hours to prepare precursor solutions with 15 concentrations of 10 wt%, 12 wt%, and 14 wt%, respectively. 16 A compensating phase modular continuous electrospinning line, consisting of a base fabric 17 unwinder, a drying device, a tension control device and a fabric winding device has been 18 developed by our group [21] . Its electrospinning main box contains multiple sets of modules 19 with mushroom-head shaped spinnerets, a sliding platform, a solution reservoir, a peristaltic 20 pump, and a high voltage direct-current power supply, as displayed in Fig. 1(a) . Separate 21 electrospinning modules could eliminate the problems of electric field interferences between 22 jets and capillary blockage, resulting in a homogeneous sub-micron fibrous membrane with 23 excellent filtration performance. Furthermore, the output speed of 0~1 m min -1 and width of 24 1~1.6 m is 150 times that of traditional single-needle electrospining; this is an important step 25 toward industrializing sub-micron fiber production. 26 14 wt%, 12 wt%, and 10 wt% PAN solutions were separately placed at one of the modules, 27 and homogeneous sub-micron fibrous membranes of with different areal densities (0.2, 0.5, 1.0, 28 1.5, and 2.0 g m -2 ) were produced by regulating the feeding speed of the substrate fabric. The 29 voltage for each module was 55 kV, and the supply rates for all solutions were the same at 80 30 ml h -1 . The electrospun fibers were deposited on the surface of a traversing substrate 18 cm 31 away from the spinneret. The slide platform moved back and forth at a speed of 9 m min -1 , 32 maintaining the uniformity of the prepared fibrous webs. Following deposition, the membrane 33 was transferred to the drying device to vaporize off the residual solvent. 34 The electrospun fibrous membranes with hierarchical structure were abbreviated as PAN- 35 x/y/z. Accordingly, three types were fabricated: (a) PAN-10/12/12, in which 12 wt% PAN 9 The hierarchical-structured PAN-10/12/14 sub-micron fiber membrane was placed between 10 two layers of PE/PP hot-air nonwoven fabrics chosen as the substrate layer (detail explanation 11 seen in Supplementary information), forming a "sandwich" structure. The composite material 12 was next subjected to a hot-rolled bonding process that included point and surface bonding (PB 13 and SB) operations, as shown schematically in Fig. 1 The linear speed, temperature, and pressure of the roller in the bonding process were the 18 detailed parameters affecting the performance of composite materials, which were analyzed via 19 orthogonal experiment and then all the factors were selected to design a whole factor test 20 scheme (Table S1 ), exploring the effect of parameters on the performance of composite 21 materials and further optimizing the process parameters. 22 23 The obtained composite material was additionally subjected to the flat-net through-air 24 penetration bonding procedure in order to create a through-air thermally bonded nonwoven, as 6 1 shown in Fig. 1(c) . In addition, the impact of process parameters (through-air temperature and 2 processing duration) on the materials' properties was investigated, as shown in Table S2 . 3 The surface morphology of sub-micron fibrous membranes was characterized by a scanning 5 electron microscopy (SEM, TM3000, Hitachi Group, Japan), and fiber diameters were 6 calculated from SEM images using Nanomeasure 1.2 software. The samples' thickness and 7 weight were determined using a thickness gauge (YG141N, Nantong Hongda Instruments Co., 8 Ltd., China) and an electronic balance (MS105DU, Mettler-Toledo Group, Switzerland), 9 respectively. The pore size distribution of fibrous webs was determined by a capillary flow 35 The pristine sub-micron fibrous monolayers had a smooth surface morphology with no beads, 37 and fibers were uniformly distributed within the membrane, as shown in Fig. 2 (a-c). Fiber 38 diameters increased from 142 to 400 nm with the coefficient of variation (C.V.) ranging from 39 17.5% to 24.6%, which was attributed to the rising concentration of PAN precursor solutions 40 ( Fig. 2 (d-f)). As the solution concentrations increased, the viscosity increased too, and the 1 accompanying rise in intermolecular forces made it more difficult to draft the jets forming sub-2 micron fibers in the electrospinning process. Similarly, the pore size of fibrous membranes 3 followed the same trend as fiber diameters ( Fig. S4(a) ). This was impacted by the areal density 4 and stacking pattern of fibrous webs, further affecting the material's filtration performance. [22] 5 For fibrous membranes with the same areal density, the fiber number per unit area and packing 6 density decreased with the increase of fiber diameter, resulting in a larger average pore size. in fiber diameters between the layers in the hierarchical-structured fibrous membrane. Fig. 3 (b) 19 shows the cross-sectional image of PAN-10/14/14 sample. 20 The pore size distribution and average pore sizes of the prepared PAN sub-micron fibrous 21 membranes with various shapes but a comparable areal density of ~1.5 g m -2 were displayed in left-shift pore size distribution. PAN-10/12/14 with tri-layer structure, on the other hand, had 28 an average pore size of 1.031 µm, which was similar to that of PAN-10/12/12 sample. 29 Compounding with a PAN-10 fine fibrous layer with a smaller pore size had a substantial 30 impact on the pore size of the composite material, according to the results. S4(c)), the rate at which the resistance rose might be calculated using the formula: 16 on a fitted curve, where y is the pressure drop, and x is the areal density of y = 179x -22.8 17 the fibrous membrane. As the fiber diameters increased, the slope of the fitted line dropped, 18 indicating that the fiber diameter exhibited an apparent effect on the membrane's filtration 19 performance. As the sample's areal density grew, smaller fibers were allowed for fabrication 20 of membranes with higher fiber packing densities and smaller average pore sizes, and more 21 sub-micron fibers overlapped to form a thicker membrane per unit area. The internal aperture 22 channels were vertically enlarged, which reduced the average pore sizes along the horizontal 23 direction and improved the membrane's efficiency against sub-micron particulate matter. [23] 24 Quality factor (Qf) is a critical metric for assessing the overall performance of a filter material 25 ( Fig. S4(d) ). As the areal density of fibrous membranes rose, Qf of the material generally 26 decreased. When the membrane's areal density exceeded 1 g m -2 , Qf value of the prepared PAN- 27 10 sample declined dramatically, indicating that the rise in areal density had a minimal effect 28 on the overall filtration performance of the filter material. For PAN-14 sub-micron fibrous 29 membrane, Qf was relatively higher with the decrease of filtration efficiency. Therefore, to 30 maximize the filtration performance of composite materials, it is necessary to reduce the 31 filtration resistance as much as possible while maintaining better filtration efficiency of the 9 1 fibrous membranes. 2 In addition, the filtration performance parameters of PAN sub-micron fiber membranes with 3 various typologies and similar areal density were depicted in Fig. 3 with larger fiber diameters extended the pore size of structural aperture channels in composite, 14 decreasing airflow resistance (Fig. 3(c) ). It was also confirmed that the hierarchical structure of 15 the sub-micron fibrous membrane, including multiple layers with different fiber diameters and 16 pore sizes, had a favorable impact on enhancing the overall filtration performance. 17 18 Due to the limited mechanical strength, the electrospun fibrous web cannot be used in practical 19 applications without the addition of a nonwoven fabric substrate to protect it. The bonding 20 progress for the composite material is considerably crucial for solving the problem of easy 21 peeling between the two, especially in post-processing, of which the influence on the 22 composite's performance is worth being analyzed extensively and methodically. To improve 23 the interlaminar fracture toughness and suppress the delamination in the form of laminated 24 nonwoven fibers in practical application, the nonwoven fabric substrates and the electrospun 25 fibrous web were compounded via the hot-rolled and through-air bonding processes, of which 26 the former included PB and SB processes. 27 3.2.1 Morphology and structure 28 The surface morphology of bonded composite materials treated by a hot-rolled process is 29 displayed in Fig. 4(c, d) , with the pristine membrane as the comparison (Fig. 4(a) ). The PE/PP 30 fibers were heated and pressed until they melted and bonded together in a block using a hot- 31 rolled PB method (Fig. S6(b) ). The core layer (electrospun fibrous web) was distorted by 32 heating and pressing around the rolling points after the nonwoven layer was revealed, resulting 33 in a plate-like adhesion. As displayed in Fig. 4(b, c) , the produced rolling points were like the 34 "crystalline zone" with dense structure, while the other area was similar to the "amorphous 35 region" with fluffy spatial structure. There was a severe adhesion with large areas among fibers 36 in the fibrous membrane (Fig. 4(d) and Fig. S6(c) ), which was because the heat emitted by the 37 roll was directly transferred to the substrate in the hot-rolled SB process. The PE/PP fibers were 38 heated and pressed to quickly deform and bond with one another. When heat was transferred to 39 the core layer, the electrospun membrane was deformed under pressure and even locally 40 carbonized or adhered with the PE/PP fibers, causing the fibrous membrane to be damaged and 41 blocked in a larger area, analogue to "continuous plane"-like adhesion. To summarize, it is 42 preferable to preserve the original microstructure and surface of the material treated using PB 43 process rather than SB process in a hot-rolled process. Nonetheless, the former only causes 1 structural damage to the fibrous membrane at the nip points, resulting in the formation of a local 2 bond point, whose size is determined by the size of the engraved roll nip point. Fig. 4(h-j) , along with a pristine specimen for comparison. For the 19 pristine specimen, there was an obvious delamination between the substrate fabric and the 20 electrospun fibrous core-layer (Fig. 4(h) ). Furthermore, the heat melt bonding of PE/PP fibrous 21 surface layer with the sub-micron fibrous core layer pressed at the rolling points resulted in the 22 relatively thinner and compacted bonding points in the point-bonded composite material ( Fig. 23 4(i)). The material was fluffy at the non-rolling point with an intact structure, however, there 24 was a change in material thickness in the transition zone between the rolling point and the non- 25 rolling point. The hot-air composite material was generally fluffy as the electrospun fibrous 26 core-layer was not squeezed by the external extrusion ( Fig. 4(j) ). 1 2 Peeling strength 3 The peeling strength test was used to assess the bonding fastness between the substrate layers 4 and the electrospun fibrous web. The higher the stripping strength, the better the material's 5 integration molding effect. Fig. 5(a) depicts the stripping strength of materials prepared by a 6 hot-rolled process under different temperatures. With the same pressure (3 MPa) and feeding 7 speed (1 m min -1 ), the stripping strength of the composite material gradually increased as the 8 treatment temperature increased. When the temperature was < 120 ℃, the composite treated by 9 the SB process showed a negligible bonding effect among layers with stripping strength less 10 than 1 cN, which was due to PE/PP fibers being unable to be melted with heat at temperature 11 lower than 120 ℃. The stripping strength of the material rapidly exceeded 6 cN at a temperature 12 of 120 ℃. The bonded material produced by PB process demonstrated a stronger bonding effect 13 among the fibrous layers. Furthermore, as the heating temperature and pressure increased, the 14 material's peeling strength improved, while its overall integrated structure became more 15 prominent ( Fig. 5(b) ). 16 As shown in Fig ℃ samples (Fig. 5d(i) ). The behavior could be explained by the lack of adhesion between the 17 substrate layer and the core layer. When the temperature in the hot-rolled PB process was ≥ 110 18 ℃, there was only one fracture peak in the material's tensile curve, indicating that the layers in 19 the treated material were firmly bonded together, and that there was no stripping or 20 delamination under external force. As the treatment temperature increased, the breaking 21 strength of the hot-rolled SB bonded material increased slightly, but the elongation at break 22 decreased (Fig. 5d(ii) ). This was related to the melting and bonding of PE/PP fibers using heat 23 and pressure. There were two fracture peaks in the stretching curves, indicating that the inter-13 1 layer bonding forces were insufficient to resist the external forces, whereas the sample would 2 easily peel and delaminate during the stretching process, resulting in the different fracture. 3 The transverse tensile curves of the materials under different hot-rolled processes are shown 4 in Fig. 5d(iii, iv) . The breaking elongation of material increased as the breaking strength 5 decreased ( Fig. 5d(iii) ). Furthermore, the difference of the fracture time between layers was 6 improved. The bonding strength between the fibrous layers was clearly sufficient to resist the 7 external force during the stretching process, and no obvious delamination occurred with the 8 improved synergy. The initial modulus of the materials treated in the same way but at different 9 temperatures increased significantly (Fig. 5d(iv) ). Because of the lower elasticity of the inter- The effect of temperature and pressure on the breaking strengths of a through-air bonded 19 composite is depicted in Fig. 5(e, f) The filtration performance of the composite materials after hot-rolled bonding treatment is 3 shown in Fig. 6(a-c) . The filtration efficiencies of the composites prepared using two kinds of 4 hot-rolled processes with the same pressure and feeding speed remained constant. With the 5 increase of heat temperature, there was a slight fluctuation. However, the pressure drop 6 continued to rise, and the comprehensive filtration performance of the materials deteriorated as 7 Qf fell. The obtained surface-bonded composite encountered higher air resistance than that of 8 the point-bonded material at temperature above 100 °C, which was attributed to the larger 9 adhesion areas blocking the air passage. Furthermore, when the temperature rose above the as temperature increased. 19 In addition, the filtration performance of the composite materials treated with various 20 parameters is interpreted in Fig. 6(c) . The sample' filtration efficiency varied slightly as the 21 heating temperature and pressing stress increased. On the other hand, there was a noticeable 22 increase in pressure drop and a decrease in Qf of the composites, which could be attributed to 23 the melting and bonding together of the substrate fabric at the rolling points, which completely 24 blocked the electrospun fibrous membrane in these locations. Furthermore, as the pressure or 25 temperature increased, the molten components spread over a larger surface, increasing filtration 26 resistance even further. 27 According to the results of the above analysis, the hot-rolled PB process had a minor impact 28 on filtration efficiency but a significant impact on the pressure drop of the thermally bonded 29 material. The ranges (R) of pressure drop for the three factors: temperature, pressure, and liner 30 speed were in the order (R pre > R tem > R lin ) in the orthogonal test results, as shown in Tables S4 31 and S5. As a result, the roller pressure had a greatest influence on air resistance followed by 32 temperature, with the linear speed having the least influence. This was attributed to the small 33 linear velocity change interval in the experimental design. The linear velocity had no effect on 34 the pressure drop of the material at low speeds. However, an increase in roller pressure or 35 temperature resulted in a linear increase of the material's pressure drop (Fig. S8 ). The through-air treatment of the composite material allowed PE/PP fibers to adhere to 7 electrospun fibrous webs. This, in turn, affected the material's filtration performance. For 8 further investigation, the extent of the impact of the through-air bonding process on composite's 9 performance is represented in Fig. 6(d, e) . The filtration efficiency of the sample treated at 130 10 ℃ for 2 minutes was comparable to that of the pristine, while the pressure drop increased 11 slightly. In this case, heat was insufficient to facilitate adhesion between PE/PP fibers and sub-12 micron fibrous webs. The filtration efficiency of the material remained unchanged as the 13 temperature raised further, whereas the air resistance of the material increased significantly. 14 This was attributed to the larger adhesive areas between PE/PP fibers and sub-micron fibrous 15 webs, obstructing airflow transmission channels within the fiber membrane. Overall, the 16 filtration performance of the materials degraded as the treatment temperature increased. 17 The material's filtration efficiency showed a slight upward trend as the processing time 18 increased for the composite material treated at different time and a through-air temperature of 1 130 ℃. When the treatment time was increased from 2 to 6 minutes, the sample's resistance 2 increased significantly at first, and then slowed down, and finally stabilized due to the limited 3 diffusivity of the molten PE/PP fiber component. When PE/PP fibers completely blocked the 4 sub-micron fibrous webs, no new adhesion was formed, and the air resistance did not increase. 5 As a result, the material's Qf decreased rapidly at first, then steadily. 6 Based on the result of the above analysis, the sub-micron composite material treated at 130 ℃ Table S7 after data standardization were weighted 23 to calculate their total score under various process parameters based on the ω of each index in 24 Table S6 . The closer the total score is to 1, the better the overall performance of the material. 25 As shown in Fig. S7 , the specimen processed at a hot air temperature of 130 ℃ for 4 minutes 26 showed the highest total score, implying that the overall performance of the composite material 27 was the best. The preliminary analysis results presented above were confirmed. 28 29 by hot-rolled PB and through-air bonding processes 30 The properties of the thermally bonded materials were further compared and investigated to 31 exhaust differences between the two bonding processes. The composite material prepared by a 32 through-air bonding process with 130 °C for 4 minutes was recorded as the through-air 33 composite material, and the material prepared by hot-rolled PB process at 110 °C, pressure of 34 3 MPa, and linear speed of 1 m min -1 was marked as the point-bonded composite material. The thickness of the point-bonded composite material was less than that of the through-air 38 composite, as shown in Fig. 7(a) . It was also discovered that the two bonded composites were 39 thinner than the pristine, which corresponded to the cross-sectional morphology exhibited in The longitudinal and transverse bending stiffness of the through-air composite materials 5 differed significantly. It was especially noticeable along the longitudinal direction, where no 6 exact value could be measured. In the transverse direction, however, the value was lower (5.82 7 mN cm), indicating high flexibility (Fig. 7(a) ). In contrast, the point-bonded composite 8 material's bending stiffness in the longitudinal and transverse directions was reduced by 50% 9 compared to the pristine sample, and its softness was greatly improved, making it suitable for 10 the processing of pleated filter materials. Because of the faster shrinkage rate, the through-air Air permeability is also important for the filter media, which directly reflects the performance 3 of gas molecules passing through the fabric. As shown in Fig. S12 , pristine sample possessed a 4 better air permeability with value of 122.3 mm s -1 , while the permeability became poor after 5 fibrous membrane was bonded thermally, especially through point-bonding process. The air 6 permeability of through-air composite material grew a little worse with value of 94.7 mm s -1 , 7 and the point-bonded material showed a worst permeability (51.6 mm s -1 ). The results indicated 8 that the thermally bonding process presents a negative effect on the permeability of composite 9 materials, owing to the induced adhesion between fibers. Good permeability performance 10 ensures the low resistance of air penetration, saving energy for filter media. Generally, the 11 through-air fibrous composite material presented superior air permeability, which is beneficial 12 in practical filtration application. composite increased by 23.8%, significantly less than that of the point-bonded material (41.8%). 19 The results were consistent with the air permeability of samples discussed above. The reason 20 was that the structure was undamaged with just several adhesions between PE/PP fibers and the 21 surface of sub-micron fibrous web for the through-air composite. While, the point-bonded 22 material was completely fused and bonded together at the rolling points, of which the 23 morphology and structure were destroyed badly, possessing bad permeability, which was not 24 beneficial for the air passing through and led to the increase of pressure drop. In general, the 25 composite process displayed a negative impact on the filtration performance of the material. 26 The through-air bonded composite outperformed the hot-rolled point bonded composite in 27 terms of filtration performance while maintaining a good material integration and structural 28 stability. 29 30 The schematic diagram of the filtration mechanisms for the prepared thermally bonded 31 composite filters is illustrated in Fig. 8 . The morphology and structure of the composite 32 materials have changed after thermally bonding, as discussed above, which affects the gas flow 33 around the fibers and the capture of airborne particles, when compared to the pristine fibrous 34 material ( Fig. 8(a) ). For the point-bonded material ( Fig. 8(b) ), there are "crystalline zone" with 35 dense structure and "amorphous region" with fluffy spatial structure. The airflow cannot pass 36 through the "crystalline zone" without pores, and its streamline will turn back after hitting the 37 "wall", evidently increasing air resistance. Meanwhile, the airborne particles with different 38 sizes in the gas streamline could be deposited on the surface of the "crystalline zone" primarily 39 through sieving mechanism. The gas flow around the fibers and particle collection through 40 composite filtration mechanisms in the other region (amorphous region) with the original 41 morphology are the same as in the pristine sample. The surface-bonded fibrous material is 42 scattered with "continuous planar"-like zones with no air permeability (Fig. 8(c) ), where the 1 gas flow dashes against the surface of the zones, changing the direction of the flow fluid, 2 resulting in a sharp increase in air resistance, which improves air permeability and increases 3 gas velocity through rest-limited regions. Of course, the entrained particles are captured in the 4 "continuous planar"-like zones by the sieving mechanism. In the other regions, due to the 5 increased gas velocity, inertial impaction proves to be the more dominant means of large 6 particle capture. [24] For the through-air bonded material with an "axon-dendrites" structure 7 (Fig. 8(d) ), the value of K n is 0.005 for the formed "axon" region with a width of ~26 µm, so 8 the gas flow around it belongs to the slip-flow regime where the air molecules around the fiber 9 surface are with segmental slip flow. While the air flow around the surrounding "dendrites" 10 with sub-micron diameter is assigned to the transition regime. It has been demonstrated that the 11 drag force on the periphery of the "axon" is larger than that of the transition regime, due to the 12 direct impact between air molecules and the "axon" in the slip-flow regime, [25] which is the 13 reason for the increased pressure drop of through-air bonded material. In summary, the "sandwich" structured composite fibrous membrane was fabricated using free 1 and pressure drop of 127.4 Pa for aerosol particles with a diameter of 260 nm under airflow 2 velocity of 5.33 cm s -1 . A through-air bonded composite material with favorable morphology 3 and structure exhibited greater rigidity, better physical properties (peeling strength, mechanical 4 breaking strength), and filtration performance than a hot-rolled bonded material. The composite 5 material demonstrated the best formability, superior peeling strength, and breaking strength 6 after through-air bonding treatment at 130 ℃ for 4 min, with a filtration efficiency of 99.44% 7 and a slightly increased pressure drop. The property-enhanced composite material, which was 8 composed of hierarchical-structured sub-micron fibrous webs and nonwoven fabrics using a 9 low-energy and low-cost through-air thermally bonding process with high yield fabrication, 10 demonstrated promising commercial and industrial productions. Air pollution and human health: from local to global issues Ambient particle inhalation and the cardiovascular 24 system: potential mechanisms Direct and indirect effects of particulate matter 27 on the cardiovascular system Rational use of face masks in the COVID-30 19 pandemic Presumed asymptomatic carrier 33 transmission of COVID-19 Electrospun ultrafine fibers for advanced face masks Air filtration in the free molecular flow Regime: A review of 37 high-efficiency particulate air filters based on carbon nanotubes Removal of nanoparticles from gas streams by fibrous filters: A review Air filtration Recent advances in polymer 4 nanofibers Nanofibrous filtering media: Filtration problems and solutions from 6 tiny materials Effect of face velocity, nanofiber packing density and 8 thickness on filtration performance of filters with nanofibers coated on a substrate Preparation of hierarchical structured nano-sized/porous poly(lactic acid) 11 composite fibrous membranes for air filtration Sandwich-structured fibrous membranes 14 with low filtration resistance for effective PM2.5 capture via one-step needless electrospinning Development of extra-sonic composite geotextile Thermal oxidative degradation of bicomponent PP/PET fiber during 18 thermal bonding process Thermal bonding of 3D nonwoven shells Single process production of 3D nonwoven shell structures: Part 2 23 CFD modelling of thermal bonding process Electrospinning fabrication and performance evaluation of 26 polyacrylonitrile nanofiber for air filter applications Large-scale preparation of micro-29 gradient structured sub-micro fibrous membrane with narrow diameter distribution for high-30 efficiency air purification Polyacrylonitrile/poly(acrylic acid) nanofibrous membranes for high efficiency particulate air 34 filtration, Fibers Polym Polyacrylonitrile/polyimide 36 composite sub-micro fibrous membranes for precise filtration of PM 0.26 pollutants Aerosol filtration application using fibrous media-an industrial perspective Slip-effect functional air filter for efficient 41 purification of PM2 Formal analysis, Data curation, Writing -Original draft Hongyu Dai: Conceptualization, Validation Supervision, Funding acquisition ☒ The authors declare that they have no known competing financial interests or personal 12 relationships that could have appeared to influence the work ☐ The authors declare the following financial interests/personal relationships which may be 15 considered as potential competing interests The hierarchical-structured electrospun fibrous web is fabricated with high yield. 24 2. The fibrous filter with tortuous channels shows superior air filtration performance Through-air thermally bonded composite exhibits more rigid, better physical property Reinforced filter bonded thermally with save-energy and low cost has good prospect