Introduction

As one of the most basic but vital activities of human beings, respiration shoulders the responsibility of transferring mass and energy through the clean air to maintain the normal metabolism of the human body1. The suspended particulate matters (PMs) with very small size in the air can carry a variety of pathogenic bacteria, viruses and toxics and thus pose a serious threat to public health via the respiratory system of the individuals1,2,3. Notably, tiny PMs (aerodynamic diameter ≤0.3 μm) are highly permeable and can induce serious respiratory diseases, cardiovascular diseases and even take our lives, which have become another crime culprit to the life safety1,2,3. Considerable facts in recent years have proved the damage of polluted PMs, such as severe acute respiratory syndrome coronavirus (SARS-CoV), influenza virus, Middle East respiratory syndrome (MERS)-CoV, human rhinovirus, and respiratory syncytial virus (RSV)1,2. Particularly, the abrupt outbreak of COVID-19 triggered by fatally novel coronavirus have unexpectedly caused billions of infections and tens of millions of deaths and been threatening the living, health and influencing the daily life of all human beings1,2,4. Therefore, the protective materials and devices to effectively prevent human body from the harm of polluted air are urgently anticipated.

Fibrous nonwoven filters with the isotropical porous structure can effectively block the airborne pollutants entering the respiratory system and is the optimal candidate for body protection4,5,6,7,8,9. The melt-blown nonwovens utilized for commercial face mask always possess the large fiber diameters (>1 μm) and pore size (over 5 μm) and thus show limited PM0.3 filtration efficiency (<99%)10,11,12. Besides, they capture PM0.3 at the expense of their thickness (>90 μm) and base weight (>20 g m−2), inevitably resulting in heat-moisture discomfort and unnecessary consumption of raw materials10,11,12. By contrast, the emerging solution-electrospinning technique brings the dawn of reducing fiber size to less than 1 μm and pore size of filters to 2-5 μm2a)55. Afterwards, the submicron-fibers could be prepared by the synergistic effect of high drafting of polymer jets and solvent volatilization (Fig. 2a)55. Considering that the properties of precursor solution strongly influence the formation of fibers, the amphiphilic dodecyltrimethylammonium chloride (DTAC), composed of hydrophilic ammonium group and lipophilic alkane chain, was introduced to simultaneously tailor the conductivity, viscosity and surface tension of PA6 solution by ion-dipole interaction44,46,50. Under the effect of electrostatic field, DTAC could be assembled around PA6 molecular chains and significantly weaken molecular chain interaction and promote their electrostatic repulsion, which beneficially resulted in the splitting of dilute jets and the formation of dual-scale fibers44,46,50. Moreover, the hydrophilic DTAC of jet surface could also attract the bad solvent (water) to replace the good solvent (Hexafluoroisopropanol, HFIP) and solidify dual-scale fibers. The amphiphilic DTAC could easily dissolve and disperse in the PA6 solutions (Fig. 2a). Figure 2b presents the effect of DTAC on solution properties. Different from the pure PA6 solutions, 5.5 wt% PA6 solutions doped with different DTAC ratio presented the enhanced conductivity and the decreased surface tension and viscosity. As the doped content of DTAC increased from 0 wt% to 12 wt%, the conductivity of PA6/DTAC solutions raised from 3 to 454 μS/cm while the viscosity of PA6/DTAC solutions decreased from 166 to 138 mPa·S, and the surface tension of PA6/DTAC solutions reduced from 20.53 to 17.80 mN/m.

Fig. 2: Formation of electrospun dual-scale fibrous networks.
figure 2

a Diagram and optical image of the formation of dual-scale fibers. b Properties of 5.5 wt% PA6 solutions doped with various DTAC contents. SEM images of (c) PA6, d PA6/DTAC-1, e PA6/DTAC-2, f PA6/DTAC-3, g PA6/DTAC-4 fiber filters. Diameter distribution of (h) 5.5 wt% PA6 fibers and (i) dual-scale fibers. j Fiber diameter comparison of this work and other spinning techniques. Error bars represent the standard deviation of the measured properties and the samples are at least three. Source data are provided as a Source Data file.

Full size image

Notably, electrospun PA6 fibers produced from dilute precursor solution showed the beaded structure (Supplementary Fig. 2a–c). But the beads within the PA6 submicron-fibers gradually decreased with the increased solution concentration from 3 to 5 wt%, and they completely disappeared when the solution concentration exceeded 5.5 wt% (Fig. 2c and Supplementary Fig. 2d). The formation of eletrospun beaded fibers results from the Rayleigh instability of polymeric jets2. Then, various PA6 and PA6/DTAC fibrous mats were prepared by utilizing a multi-needle electrospinning machine (NEU, DAIEI KAGAKU SEIKI MFG. Co., Ltd., Japan). The speed of collector and reciprocating spinnerets was set to 5 m/min and 10 cm/min, respectively. The fibrous filter with the size of 30×70 cm2 could be uniformly fabricated through rotary collector and reciprocating spinnerets.

Characterizations

The solution properties (conductivity, viscosity, and surface tension) were measured by conductivity meter (FE32-Standard, Mettler-Toledo Group, Switzerland), viscometer (DV3T, Brookfield Ltd., America), and surface tension meter (DCAT11, Dataphysics Instrument Ltd., Germany), respectively. Jet splitting image was captured via high-speed camera (i-SPEED 716, Nikon, Japan). The morphologies of the electrospun fibrous mats were characterized by scanning electron microscopy (SEM; DXS-10ACKT, Hitachi Group, Japan). The diameter of the electrospun fibers was determined utilizing the Nano Measurer software. The wettability of samples was assessed by the contact angle goniometer (OCA15EC Dataphysics, Germany). The pore sizes and pore size distributions of membrane were characterized by a capillary flow porometer (POROLUXTM 100 FM, IB-FT, Germany). The base weight of mat was acquired by an electronic balance (MS105DU, Mettler-Toledo Group, Switzerland). The thickness of fibrous networks was gauged from their cross-sectional SEM images by the Nano Measurer software. The visible light transmittance of specimens was measured through a UV-Vis-NIR spectrometer (UV3600, Shimadzu Instruments Co., Ltd, Japan, with a standard barium sulfate whiteboard for the calibration) with a diffuse integrating sphere. Based on the USA standard (IEST-RP-CC52.2-2007) and the European standard (EN779: 2012) for air filters, the filtration performances of samples with effective testing area of 100 cm2 were measured under continuous airflow velocity of 32 L min−1 by an automated filtration testing machine (Model 8130, TSI Group, America). The instrument could produce charge-neutralized monodisperse solid NaCl particles with a mass median diameter of 260 nm and count median diameter of 75 nm. Each sample was tested for three times at different areas and then averaged to obtain final filtration results. The filtration efficiency (η) was automatically calculated by the commercial machine according to the equation (\({{{{{\rm{\eta }}}}}}=1-\frac{{{{{{\rm{C}}}}}}}{{{{{{{\rm{C}}}}}}}_{0}}\)), where C and C0 respectively represent the aerosol particle number concentrations at the testing outlet and the testing inlet. Quality factor (QF, Pa−1) was calculated by the equation (\({{{{{\rm{QF}}}}}}=-\frac{{{{{\mathrm{ln}}}}}(1-{{{{{\rm{\eta }}}}}})}{\Delta {{{{{\rm{P}}}}}}}\)), where ΔP is the pressure drop. The porosity (P) was calculated by the equation (\({{{{{\rm{P}}}}}}=1-\frac{{{{{{\rm{w}}}}}}}{{{{{{\rm{d}}}}}}{{{{{\rm{\rho }}}}}}}\)), where w, d and ρ respectively represent the base weight, thickness and density of the membrane. The long-term filtration process of the filter was conducted by the self-made filtration apparatus (continuous particle airflow containing over 100000 particles, Supplementary Fig. 7) and its filtration properties was then tested by commercial instrument. The surface potential of membrane was conducted by an electrostatic field tester (EFM 023, Kleinwaechter GmbH, Germany). The water vapor transmittance rate of filters was performed with a fabric moisture permeability meter (FX3081-CM15, TEXTEST, Switzerland) according to the standard of GB/T 12704.1–2009. The air permeability test was determined by a digital fabric air permeability meter (YG461E, Ningbo Textile Instrument Factory, China) according to the standard of GB/T 5453-1997. The infrared transmittance was obtained by the Fourier transform infrared (FTIR) spectrometer FTIR, Nicolet 8700, Thermo Fisher, USA) equipped with a diffuse gold integrating sphere (PIKE Technologies). The thermal images of filters worn by a volunteer (male, 28 years old) were recorded by an infrared camera (HM-TPK20-3AQF/W, Hangzhou Hikvision Electronics Co., Ltd., China). Based on the Chinese standard (GB/T 20944.3-2008), colony-counting method was utilized to quantitatively evaluate the antibacterial activity of the prepared filter57. The pure PA6 submicron-fiber membrane and PA6/DTAC dual-scale fibrous membrane were separately chosen as blank and experimental samples. And their antibacterial durability was also evaluated after six months of storage.The antibacterial rate (α) could be calculated by the equation (\({{{{{\rm{\alpha }}}}}}=-\frac{{{{{{\rm{c}}}}}}1}{{{{{{\rm{c}}}}}}2}\)), where c1 and c2 represent the live bacterial concentration of blank and experimental samples. The live bacterial concentration is the product of colony count and dilution times of the bacterial culture solution. Error bars represent the standard deviation of the measured properties and are calculated from at least three samples.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.