Figure 2a presents optical photographs of the solution during the preparation of CS@Ag core-shell fibers. The AgNPs solution should be initially yellow, then brown, and finally black (Mohammadalinejhad et al., 2019; Wang et al., 2016). However, the solution has been clarified during the experiment. only the color of the CS fiber changed as described above. With continuous reduction of Ag+, the amino and hydroxyl groups on the surface of CS fibers acted as anchors for AgNPs via the atom-by-atom growth, and finally CS fibers were in situ coated with reduced AgNPs (Huo et al., 2020; Shen and Feng, 2018). To verify the amount of the reduced AgNPs that were adhered to the fibers, the CS fibers were weighed before and after the treatment. The utilization rate of AgNPs was calculated by the following formula:

Where m1,m2, m3 refer to the mass of CS fibers, AgNO3, and CS@Ag core-shell fibers, respectively (MAg = 108. MAgNO3 = 170). As shown in Figure 2b, the utilization rate of AgNPs is between 92-97%, indicating that 92-97% of the reduced AgNPs adhered to the fibers. The remaining 3-8% of AgNPs should be silver mirrors (figure 2a) attached to the inner wall of the beaker.
To determine whether the reduction product is elemental silver or not, the CS@Ag core-shell fibers were analyzed by X-ray diffraction (XRD) analysis. It can be seen from Figure 2c that there are four main peaks at 2θ values of 38.5°, 44.6°, 64.8°, and 77.7°. These peaks correspond to (1,1,1), (2,0,0), (2, 2,0), (3,1,1) diffraction planes of Ag crystals (Ming et al., 2019; Parida et al., 2020; Shen and Feng, 2018), which further corroborates that the surface of the CS fiber is indeed AgNPs. With the increase of AgNO3 concentration, the diffraction peak at 38.56° becomes more intense, and the half-value width decreases from 0.697° (CS@Ag-12) to 0.468° (CS@Ag-36). This phenomenon shows that the crystalline grain size of AgNPs gradually increases and the crystals develop more intact with the increase of AgNO3 concentration (Wu et al., 2015; Ye et al., 2019).
Figure 2d is the thermogravimetric analysis (TGA) curves of the CS@Ag core-shell fibers with different AgNPs loadings. CS displays three stages of thermal weight loss. The first stage corresponds to the volatilization of water. The second stage is related to the breakage of the CS molecular skeleton unit. And the third stage stems from the carbonization of CS (Bazmandeh et al., 2020; Chen et al., 2020). After CS fibers were coated with AgNPs, the thermal weight loss of CS in the first stage is significantly slowed down. Here we boldly guess that it was because the AgNPs evenly encapsulate the CS fibers so that the water volatilization was drastically reduced and the thermal stability was improved. In addition, it can be seen from the remaining amount that the number of AgNPs adhering to the fiber gradually increases with the increase of AgNO3 concentration.
To verify the above-mentioned conjecture of the TGA results, scanning electron microscopy (SEM) and energy dispersive spectroscopy (EDS) were used to characterize the surface morphology and element of pure CS and CS@Ag core-shell fibers with different AgNPs loadings. Figure 3a1~a2 show the SEM images of CS fiber, the surface of which is smooth and free of impurities. Figure 3b1~b2, 3c1~c2, 3d1~d2, 3e1~e2 show the SEM images of CS@Ag-12, CS@Ag-21, CS@Ag-29 and CS@Ag-36 fibers, respectively. In the SEM images, it can be seen that AgNPs grow uniformly on the surface of the CS fiber until totally wrap it, which confirms the above conjecture. When the concentration of AgNO3 is low, the morphology of CS fibers is vaguely visible (figure 3b2) because the loading of AgNPs is less and the size of AgNPs is small at this time. As the concentration of AgNO3 increases, the loading and particle size of AgNPs increases, completely encapsulating CS fibers to form a core-shell structure. The universal meter was used to make a rough measurement of the resistance of the CS@Ag core-shell fibers. Undoubtedly, the resistance of CS fibers is infinite (Figure 3a3) because CS is not conductive. Interestingly, the resistances of CS@Ag-12, CS@Ag-21, CS@Ag-29 and CS@Ag-36 core-shell fibers are 40.9 Ω (Figure 3b3), 6.1 Ω (Figure 3c3), 0.3 Ω (Figure 3d3), 0.1 Ω (Figure 3e3), respectively. Conductive AgNPs formed a continuous phase on the surface of CS fibers, making the CS fibers conductive. And with the increase of AgNO3 concentration, the core-shell structure was improved, so the resistance of CS@Ag core-shell fibers gradually decreased. The EDS scanning test of carbon (C), oxygen (O), nitrogen (N), and silver (Ag) was carried to investigate the composition at the surface of CS@Ag core-shell fibers (figures 3e4~e6). The C, O, and N elements are originated from the long carbon chain of CS and the Ag element is derived from AgNPs (Chen et al., 2020). Based on the results of elemental mappings, it can be seen that the above Ag elements are evenly and tightly distributed on the fiber surface, and the total mass is as high as 87.2 wt%. The EDS results once again proved the formation of the core-shell structure.
The prepared CS@Ag core-shell fibers were combined with plant fibers to make functional CS@Ag/plant fiber membranes through a simple suction filtration method. To predict whether the CS@Ag/plant fiber membranes can be applied to the filter layer of masks, some key indicators related to the quality of the face mask, including air permeability, removal efficiency, and hydrophobicity, were tested. Air permeability is closely related to wearing comfort. Figure 4a exhibits the air permeability of commercial PP face mask, pure plant fiber membrane, and CS@Ag/plant fiber membranes. As the "binder" between CS@Ag core-shell fibers, the main component of plant fibers is cellulose. The strong hydrogen bonds between cellulose can make the fiber membrane tightly bonded and reduce air voids, which can be seen from the low air permeability of the pure plant fiber membrane (only 27.5 mm·s-1) (Wang et al., 2018; Wang et al., 2021; Zhou et al., 2021). Although the addition of CS@Ag core-shell fibers would break the hydrogen bonds between plant fibers, the density of CS@Ag core-shell fibers increased with the addition of AgNPs. When the same quality of CS@Ag core-shell fibers was added, the number of CS@Ag core-shell fibers will decrease. Eventually, the structure of CS@Ag/plant fiber membrane changed from loose to relatively compact. The air permeability of the sample is reduced from 349.7 mm·s-1 (S1) to 260.2 mm·s-1 (S4). Compared with the commercial PP face mask with an air permeability of 167.4 mm·s-1, the CS@Ag/plant fiber membrane are much more breathable.
If people are upon long-term exposure to PM2.5, PM2.5 particles will permeate through human respiratory tract and cause harm to them (Liu et al., 2020a). Therefore, the face mask with perfect PM2.5 removal efficiency plays an important role in human body safety protection (Liao et al., 2021a). Figure 4b shows the PM2.5 removal efficiency of commercial PP face mask, pure plant fiber membrane, and CS@Ag/plant fiber membrane. PM2.5 removal efficiency has a certain correlation with air permeability and air void size. The PM2.5 removal efficiency of the pure plant fiber membrane itself is better, with a removal efficiency of 98.7%. Therefore, the addition of pure plant fiber will effectively improve the PM2.5 removal efficiency of CS@Ag/plant fiber membrane. With the increase of the AgNPs loading on the CS@Ag core-shell fiber, the air void size and air permeability decrease, and the removal efficiency of PM2.5 gradually increased from 86.7% (S0) to 94.8% (S4). The PM2.5 removal efficiency of CS@Ag-36/plant fiber membrane surpasses the performance of commercial PP masks with a PM2.5 removal efficiency of 94.2%.
Since the COVID-19 spread through respiratory droplets, the hydrophobicity of filter layer for masks is a contributing factor to their service life and antibacterial activity (Kumar et al., 2020). To confirm whether C13H13F17O3Si successfully modified fiber membranes, hydrophilicity test for the CS@Ag/plant fiber membrane before and after treatment were carried out, and the results are summarized in Figure 4c and 4d. As shown in Figure 4c, the treated membrane has excellent hydrophobicity, while the untreated membrane is hydrophilic, and the water droplets are immediately absorbed by it. Figure 4d summarizes the water contact angle data for pure plant fiber membrane and CS@Ag/plant fiber membrane with different AgNPs loadings. The contact angles of treated samples were all greater than 140°, which indicated that the fiber membranes treated with C13H13F17O3Si changed from a hydrophilic material to a high hydrophobic material. The hydrophobicity of the fiber membranes endows itself with certain self-cleaning and antifouling properties (Wu et al., 2016).
Besides the satisfactory PM2.5 removal efficiency and air permeability, the CS@Ag/plant fiber membrane has distinctive heat dissipation function. A laser thermal conductivity meter was used to measure the in-plane and through-plane thermal conductivity of samples, and the results are summarized in Figures 5a and b. The in-plane thermal conductivity of the pure plant fiber membrane is 0.30 W·m-1·K-1, and the through-plane thermal conductivity is 0.11 W·m-1·K-1. The addition of CS@Ag core-shell fiber makes the thermal conductivity of CS@Ag/plant fiber membrane increase immediately. And with the increase of the AgNPs loading on the CS fiber, the thermal conductivity of CS@Ag/plant fiber membrane continues to increase. When the AgNPs loading on CS fibers is 36 vol%, the in-plane thermal conductivity of CS@Ag/plant fiber membranes increases to 0.89 W·m-1·K-1. Compared with commercial PP masks with in-plane thermal conductivity of 0.20 W·m-1·K-1, the in-plane thermal conductivity of CS@Ag/plant fiber membranes is 4.45 times that of commercial PP masks, showing excellent heat dissipation performance. The thermal conductivity of the plant fiber itself is low, and the heat transfer rate in the plant fiber is slow, so the connection of the plant fiber is not the reason for the rapid increase of the in-plane thermal conductivity. From the surface SEM images of samples S0 and S4 in figure 5c, the reason for the increase in in-plane thermal conductivity can be seen. The thermal conductivity of AgNPs is as high as 429 W·m-1·K-1, so the heat transfer rate is fast on the surface of CS@Ag core-shell fiber. And the heat transfer rate increases with the increase of AgNPs loading. Therefore, the connection of CS@Ag core-shell fibers is the reason for the rapid increase in in-plane thermal conductivity. To satisfy the breathability of the fiber membrane when used as a mask, the cross-section of the fiber membranes entrains a large amount of air voids. The thermal conductivity of air is extremely low, only 0.023 W·m-1·K-1, which results in a low through-plane thermal conductivity (Hu et al., 2021). Even though the loading of AgNPs on CS fibers is as high as 36%, the through-plane thermal conductivity of the CS@Ag/plant fiber membranes is only 0.15 W·m-1·K-1, which is 1.15 times that of commercial PP masks (0.13 W·m-1·K-1).
To characterize the thermal management capability of plant fiber membranes and CS@Ag/plant fiber membranes, the surface temperature of the fiber membranes during heating was measured by an infrared thermal imager every 15 s (Zhang et al., 2019). The results are shown in Figure 5d~e. At the first second, the surface temperatures of S0, S1, S2, S3, and S4 are 18.1, 18.4, 18.6, 18.9, and 23.3°C, respectively. After heating for 90 s, the surface temperature of S0, S1, S2, S3, and S4 increased to 36.6, 37.9, 44.4, 50.1, and 61.3 °C, respectively. As the loading of AgNPs increases, the corresponding surface temperature of each sample increases more rapidly. These results are consistent with the experimental results shown in Figure 5a~b. The sample S4 was made into a mask and worn by the model at room temperature (25°C). As shown in Figure 5f, the surface temperature of the self-made fiber mask is quite different from that of the commercial PP face mask after wearing it for 1 min. The average surface temperature of the self-made fiber mask increased from 24.3 °C to 27.6 °C, while the average surface temperature of the commercial PP face mask only increased from 24.6 °C to 25.7 °C. In contrast, self-made masks have a significant cooling effect by transmitting heat generated by the human body, while commercial PP masks block a large portion.
The connection of CS@Ag core-shell fibers not only endows CS@Ag/plant fiber membranes with good thermal conductivity but also low electrical resistance and high electrical conductivity (Lv et al., 2020). The resistance of CS@Ag/plant fiber membranes was measured by voltammetry. The principle of voltammetry is to measure the corresponding current by applying different voltages to the CS@Ag/plant fiber membrane. Then, the obtained data points are linearly fitted to obtain the resistance of the sample, as shown in Figure 6a. The resistances of samples S1, S2, S3, and S4 are 48.6, 37.4, 1.4, and 1.3 Ω, respectively. Electrical conductivity (σ) is calculated using the following equation (Lim and Kitagawa, 2020; Xie et al., 2020):

where L, R, and S are the sample length (m), resistance, and electrode area (m2), respectively. The calculated conductivity of samples S1, S2, S3, and S4 are 55.1, 73.7, 2228.2, and 2236.1 S·m-1, respectively. CS@Ag core-shell fibers form a conductive network in the fiber membrane. As the loading of AgNPs increases, the resistance of the CS@Ag/plant fiber membrane decreases and the electrical conductivity becomes better.
The low resistance and high conductivity of CS@Ag/plant fiber membrane make them excellent candidates for use as an electrothermal material. In a low-temperature simulation room at -11 °C, a portable power bank with an output voltage of 5 V and an output current of 2 A was used to power the CS@Ag-36/plant fiber membrane (sample S4), as shown in Figure 6c. The surface temperature of the fiber membrane was observed by an infrared thermal imager at a certain interval. The surface temperatures and the corresponding infrared thermal image are shown in figure 6d~e. In the low-temperature simulation chamber, the surface temperature of the membrane is only 3 °C. After energizing, the surface temperature of the film rises immediately and reaches the maximum temperature of 30.2 °C when energized for 10 min. Subsequently, the temperature dropped slightly. After energizing for a long time, the temperature is maintained at 21.4 °C, which is 18.4°C higher than the surface temperature of the membrane without electric heating. The CS@Ag/plant fiber membranes can provide heat to the human body in a cold environment.
As shown in Figure 6f, the power bank was used to power CS@Ag/plant fiber membrane at room temperature (25 °C). After 1 min, the temperature of CS@Ag/plant fiber membrane can reach above 80 °C (figure 6g). Since COVID-19 can be inactivated within 30 min at 56 °C (Chin et al., 2020), our findings confirm that CS@Ag/plant fiber membrane can inactivate COVID-19 with benefits of no added costs, chemicals, or work time. When people go out, CS@Ag/plant fiber membrane can not only block viruses and particles but also dissipate heat in a hot environment and keep warm in a cold environment. When people get home, it can be energized to sterilize and inactivate viruses as shown in Figure 6h.
When tenacious pathogens accumulate on the mask, improper use and disposal of the face mask may cause the risk of secondary infection to the wearer (Mahat et al., 2021). Therefore, antibacterial properties are also one of the important properties of masks. E. coli and S. aureus are typical representatives of Gram-negative bacteria and Gram-positive bacteria respectively and are also the most common bacteria in life (Liu et al., 2020b). We examined the antibacterial properties of fiber membranes against E. coli and S. aureus. Figure 7a is optical images of bacterial culture plates, showing the antibacterial activity of each fiber membrane against E. coli and S. aureus at the same dilution concentration. After 24 h of culture, the bacteria in the bacterial culture dish of the plant fiber membrane were metabolized vigorously. In contrast, the number of two kinds of colonies in the bacterial culture dish of CS@Ag/plant fiber membrane was significantly reduced. The quantitative analysis results of the antibacterial properties of the fibrous membranes (Figure 7b) show that the antibacterial rates of the plant fibers membrane against E. coli and S. aureus are only 30.99% and 20.82%. The antibacterial rate of CS@Ag/plant fiber membrane against E. coli and S. aureus is above 99%. This is because CS@Ag core-shell fibers in the fibrous membrane will release Ag+ when they come into contact with bacteria, and Ag+ has excellent antibacterial properties. The antibacterial mechanism diagram is shown in Figure 7c. The cell membrane surfaces of both Gram-negative bacteria and positive bacteria are negatively charged, and Ag+ are positively charged. So, Ag+ is easily electrostatically adsorbed on the surface of the cell membrane, leading to the destruction of the cell surface balance and the leakage of intracellular substances. In addition, Ag+ can interact with thiol groups in proteins to promote the production of active oxygen, which will damage the protein and DNA and eventually lead to the death of bacteria (Liu et al., 2021; Mohammadalinejhad et al., 2019). The antibacterial experiment results prove that CS@Ag/plant fiber membranes have excellent antibacterial properties, and can be used as a mask filter to filter inhaled or exhaled bacteria, reducing the spread of bacteria.