A novel strategy for producing high‐performance continuous regenerated fibers with wool‐like structure

Abstract

We proposed a novel approach to prepare high‐performance continuous regenerated keratin fibers with wool‐like structure by using the cortical cells and linear keratin from wool waste as reinforcement and adhesive, respectively. The spindle‐shaped cortical cells were taken from wool waste based on the different responses of cortical cells and mesenchyme in wool to the treatments of H₂O₂ oxidation and ultrasonication. The linear keratin was yielded through dissolving wool waste in the green solution consisting of starch derived dithiothreitol and protein denaturant sodium dodecyl sulfate. The recycled keratin fibers were produced by wet‐spinning of the mixture solution comprising of cortical cells, linear keratin and toughener poly(ethylene glycol) diacrylate, and crosslinked by glutaraldehyde and 4,4′‐methylenebis‐(phenyl isocyanate). The cortical cells were aligned along the regenerated fibers axis and retained quite a few α‐helical crystals of the intermediate filaments, benefitting improvement of mechanical properties. Consequently, the valuable chemical compositions and hierarchical microstructures of wool were largely inherited. Their mechanical properties, thermal stability, dyeing property, moisture absorption capability, and antistatic resistance resembled those of wool. The regenerated fibers contained 93.3 wt.% components of wool, and the amount of synthetic chemicals in the regenerated fibers was controlled to as low as 6.7 wt.%.

INTRODUCTION

A great deal of wool waste are generated from textile and slaughtering industries every year,^1–4 and they are usually used for low-value applications such as feed additives, cosmetic ingredients, wastewater treatment agents, and biomaterialsh.^1,5–9 As natural wool consists of large amount of valuable keratin (>90%),^1,10 however, extraction of keratin from wool waste and preparation of regenerated keratin materials would be more meaningful.^3,11–21

Recently, the authors proposed a green solution consisting of 1,4-dithiothreitol (DTT) and sodium dodecyl sulfate (SDS), which can completely dissolve wool waste into high concentration keratin solution for direct wet spinning, without damaging the protein backbones like conventional methods (incl. reduction,^{1,6,8,19,21,22} oxidation,^9,10,23–25 alkaline hydrolysis,^10,23,24 and ionic liquids^{10,23,24,26,27}). With the aid of a few toughening agent and crosslinkers, continuous keratin fibers (fraction of keratin = 86 wt.%) were obtained. Up to now, there are only a few researches focusing on the preparation of pure keratin fibers.^3,28–30 Moreover, the regenerated fibers are usually brittle. To improve the spinnability of the keratin solution, poly(vinyl alcohol) (PVA),^13,31 cellulose,¹² and cellulose acetate¹¹ were added in the previously reported works. However, the ultimate fibers suffer from the weakness of low fractions of keratin (e.g., 15-45 wt.%), and poor mechanical properties in comparison with natural wool.^11–14

Here in this work, the authors plan to produce continuous keratin fibers in another way after carefully surveying the microstructure of wool. It is known that natural wool fibers are covered by cuticle layers (10 wt.%),³² while their trunks, accounting for 90 wt.% of the wool fibers,^32–34 are made up of tightly structured cortical cells and amorphous mesenchyme.^32,35 Meantime, the cortical cells consist of many microfibrils,^32,36 which are assembled by smaller intermediate filaments containing lots of α-helix crystals.^24,32,37 According to these findings, a new recycling technique based on wet spinning is worked out as follows (Figure 1A). First, the spindle-shaped cortical cells are extracted from discontinuous wool waste after oxidation (Figure 1B) and ultrasonication by taking advantage of the structural difference between the cortical cells and mesenchyme. Then, the extracted cortical cells are well mixed with the purified linear keratin to prepare the solution (Figure 1C) for wet spinning, while the keratin is yielded in advance by dissolving wool waste in terms of the above-mentioned DTT/SDS system (Figure 1C). For purposes of improving toughness of the regenerated keratin fibers, a few flexible poly(ethylene glycol) diacrylate (PEGDA) are incorporated as toughening agent, whose vinyl groups can react with sulfhydryl groups of the decrosslinked keratin using 1,8-diazabicyclo[5.4.0]undec-7-ene (DBU) as catalyst^38–41 playing the role of precrosslinking (Figure 1D) and increasing intermolecular interactions. Last, the as-spun regenerated keratin fibers, in which the cortical cells serve as the reinforcement and the linear keratin as the binder, are further crosslinked for strengthening to meet the application requirements for mechanical properties. In this case, glutaraldehyde (GA) (Figure 1E) and 4,4′-methylenebis-(phenyl isocyanate) (MDI) (Figure 1F) act as the crosslinking agents through the reactions with the amino and hydroxyl groups of keratin. Our previous work³⁷ indicated that both the two kinds of crosslinking agents can effectively raise the strength of keratin fibers. The Schiff base and carbamate created after the crosslinking appear in many biomaterials and prove to be safe.^7,42–44

[IMAGE OMITTED. SEE PDF]

It is hoped that the resulting regenerated keratin fibers would not only possess wool-like structure but also higher keratin content (or lower synthetic substances content) in comparison with the earlier version.³⁷ Hereinafter, the preparation and structure properties relationships of the regenerated continuous keratin fibers are discussed.

EXPERIMENTAL SECTION

Materials

PEGDA (M_n ≈ 600), 1,8-diazabicyclo[5.4.0]undec-7-ene (DBU), GA, and MDI were purchased from Aladdin Biochemical Technology Co., China. DTT and acid blue 80 were purchased from Energy Chemical Technology Co., Ltd., China. Hydrogen peroxide (H₂O₂), hydrochloric acid, formic acid, acetic acid, acetone, ammonium chloride, ethanol, and sodium sulfate were supplied by Guangzhou Chemical Reagent Factory, China. SDS was purchased from Sangon Biotechnology Co., Ltd., China. Degreased short coarse wool waste (fiber length ≈ 10–50 mm, fiber diameter ≈ 50 μm, Figure 1G) was purchased from Hualin Wool Products Co., China, which were washed and dried at 100°C for 2 h before use.

Removal of the cuticle layers from wool waste

A total of 2.0 g wool waste was cut into 5 mm long segments and added into 100 mL formic acid at 100°C for 10 min. Then, the mixture of the swollen discontinuous wool waste and formic acid was transferred into a 150 mL jar. Cooled in ice-water bath, the cuticle layers were removed from the wool waste by ultrasonic treatment (300 W, 20 kHz) for 0.5 h in accordance with the following preset program. That is, the machine operated for 3 s and rested for 1 s. Eventually, the mixture was filtered by a stainless standard sieve (100 mesh). After washing by pure water for three times and then drying at 100°C, the cuticle layers-free wool fibers were available.

Extraction of the cortical cells

A total of 2.0 g cuticle layers-free wool was immersed in 100 mL H₂O₂ solution at room temperature. After oxidation (Figure 1B), the wool was washed and dried. Then, the oxidized wool was added into 100 mL formic acid at 100°C for 10 min. The mixture was also transferred into a 150 mL jar and cooled in ice-water bath under ultrasonication (600 W, 20 kHz) for 0.5 h following the same program as mentioned above. Afterwards, the mixture was successively filtered by two pieces of standard stainless sieves (100 and 1000 mesh). The extracted cortical cells remaining on the 1000 mesh sieve were washed by ethanol for three times and freeze-dried. To study the influence of H₂O₂ concentration and oxidation time, 1, 3, and 5 wt.% H₂O₂ solutions were used, and the oxidation times were set at 6, 12 and 18 h.

Preparation of keratin solution

Keratin was extracted following the technical path described in our previous work (Figure 1C).³⁷ A total of 20.0 g wool waste was immersed in 80.0 g aqueous solution containing 5.0 g DTT and 10.0 g SDS at a weight ratio of wool/solution = 1:4 under 80°C, and the mixture was continuously and rapidly stirred for 12 h to obtain the completely dissolved keratin solution. Next, the keratin solution was added into 1.0 L 30 wt.% NH₄Cl solution, and the pH was adjusted to 4.0 with HCl. The precipitate of linear keratin was available by tuning salting out and isoelectric point. After filtration, washing and freeze-drying, the powdered linear keratin was obtained. To prepare 20 wt.% keratin solution, 10.0 g linear keratin powder, 5.0 g SDS, and 35.0 g water were added into a conical flask and the mixture was stirred until full dissolution.

Manufacturing of the as-spun regenerated keratin fibers containing PEGDA

Two types of mixture of the 20 wt.% keratin solution and PEGDA were prepared at the mass ratios of 95:1 and 45:1, respectively, while PEGDA accounted for 5 and 10 wt.% of the total mass of PEGDA and keratin. DBU accounting for 0.1 wt.% of the total mass of the mixtures was added at the same time. After stirring and reacting for 12 h, the mixtures were centrifuged under 1000 rpm for 10 min to eliminate air bubbles. A 30G flat needle (inner diameter = 0.16 mm) served as the spinneret. The spun solution was squeezed out of the cylinder spinner under room temperature and entered a 1 m-long coagulation bath full of 30 wt.% NH₄Cl, the pH value of which was tuned to be 4.0 using HCl. By adjusting the salting-out and isoelectric point, the spinning solution was immediately precipitated in the coagulation bath and then the precipitate was stretched at a draft ratio of 2 to produce the desired as-spun keratin fiber.

Manufacturing of the as-spun regenerated keratin fibers containing PEGDA and cortical cells

Three types of mixture of the keratin solution and PEGDA were prepared at the mass ratios of 75:1, 55:1, and 35:1, respectively. DBU accounting for 0.1 wt.% of the total mass of the mixtures was also added. After stirring and reacting for 12 h, cortical cells weighting 4, 8, and 12 times of PEGDA were added to the mixtures, respectively. Eventually, the mass ratios of cortical cells, PEGDA, and keratin in the mixtures were 20:5:75, 40:5:55, 60:5:35. The as-spun keratin fibers containing both PEGDA and cortical cells were prepared by means of the same wet-spinning approach as described in the above.

Crosslinking of the as-spun regenerated fibers using GA and MDI

The pH value of the aqueous solution containing 5 wt.% GA was adjusted to 4.0 with HCl. Then, the as-spun keratin fibers were immersed in the solution for 5 min. Afterwards, the crosslinked fibers were washed with distilled water and freeze-dried. Finally, the GA crosslinked fibers were put into redistilled acetone solution containing 2.0 wt.% MDI for 1 h at 45°C and washed with acetone for three times and dried.

Characterization

Morphologies of the cuticle layers-free wool, cortical cells, and keratin fibers were examined by a HITACHI model S-4800 scanning electron microscopy (SEM). The samples were sputter coated with platinum in advance and observed at an accelerating voltage of 10 kV.

The crystalline structures of the wool waste, cortical cells, and keratin fibers were measured by a D8 ADVANCE X-ray diffractometer (WAXD, Cu K_α radiation, λ = 0.154 nm, 2θ scan range = 5°–80°, 10°/min, Bruker Corporation, Germany). The crystallinity index, CI, was estimated from^{3,26,28,45,46}: 1 $\begin{equation} \textit{CI}=\frac{{I}_{{9}^{\circ}}-{I}_{{14}^{\circ}}}{{I}_{{9}^{\circ}}} \end{equation}$where I_9° and I_14° represent the intensities at 2θ = 9° and 14°.

Tensile properties of the regenerated fibers were tested at 25°C on the SANS CMT 6000 universal testing machine according to the ISO 11566 standard. The test was performed with a gauge length of 25 mm and a crosshead speed of 5 mm/min. At least 50 specimens were measured per batch.

The molecular structure of the regenerated fibers was analyzed by Fourier transform infrared (FTIR) spectroscope (Bruker EQUNINOX 55 Spectrometer, Germany) within the wavenumber range of 4000—400/cm with 64 scans and a resolution of 2/cm. Raman spectroscopy (Thermo Fisher Scientific FT-Raman Spectrometer, America) was applied to determinate the changes in functional groups using 10 scans with a resolution of 8/cm. Solid ¹³C nuclear magnetic resonance (¹³C NMR) spectroscopy was recorded without any solvent at 400 MHz on a Bruker Avance-400 spectrometer using a 4 mm rotor to study the secondary protein conformation according to the previous works.^26,27,47 Elemental analysis was carried out on an Elementar Vario EI instrument to determine the content of N, which can be used for calculating the content of PEGDA in the regenerated fibers.

Thermogravimetric analysis (TGA) was conducted on a TA Instruments TGA Q50 at a heating rate of 5°C/min under a constant nitrogen flow.

The orientation degree, F, of the fibers was determined by an Empyrean X-ray diffractometer (Cu K_α radiation, λ = 0.154 nm, 2θ = 9° or 20°, phi scan range = 0°–180°, PANalytical B.V., Netherlands) according to: 2 $\begin{equation} F\hspace*{0.28em}=\hspace*{0.28em}1-\frac{{W}_{1/2}}{180} \end{equation}$where W_1/2 is the half-width of the diffraction peak.

The swelling ratio, Q, of the fibers was measured after soaking in 10 wt.% SDS solution for 48 h, and calculated from: 3 $\begin{equation} Q\hspace*{0.28em}=\frac{{m}_{0}/{\rho}_{p}+\left({m}^{\prime}-{m}_{0}\right)/{\rho}_{s}\hspace*{0.28em}}{{m}_{0}/{\rho}_{p}\hspace*{0.28em}}\hspace*{0.28em} \end{equation}$where m_o and m’ are the weights of the fibers before and after swelling, while ρ_p and ρ_s denote the densities of the fibers and SDS solution.

The standard moisture recovery rate, R, was calculated from: 4 $\begin{equation} R=\frac{m-A{m}_{0}}{A{m}_{0}}\hspace*{0.28em}\ensuremath{\times{}}\hspace*{0.28em}100\% \end{equation}$where m is the weight of the fibers treated at 20°C under 65% relative humidity for 24 h, while m_o is the weight of the fibers dried at 105°C to constant weight. The correction factor, A, is 0.996 at an ambient temperature of 28°C under 75% relative humidity.

The weights of the regenerated fibers before and after crosslinking were measured, and the content of the crosslinking agents (GA and MDI) that had reacted with the fibers was expressed by the increase in weight after crosslinking in comparison to that before crosslinking.

Electrical resistance was measured by the Smart Sensor AR3127 to understand the antistatic property of the fibers. Accordingly, volume resistivity was estimated from: 5 $\begin{equation} {\rho}_{v}=\frac{\textit{RS}}{L} \end{equation}$6 $\begin{equation} S=\frac{m}{\rho L} \end{equation}$where R and L are the resistance and length of the fibers, while m and ρ are the weight and density of the fibers.

To evaluate the dyeing performance according to GB/T 2378-2012 standard, 1% (owf) acid blue 80 with a bath ratio of 1:5 was used as the dye. The keratin fibers were stained at 95°C for 0.5 h, and then the absorbances of the dye solution before, A_o, and after, A, the dyeing process were measured by a Perkin Elmer Lambda 750 UV-Vis-NIR spectrophotometer. Accordingly, the dye-uptake, E, was calculated from^48,49 7 $\begin{equation} E=\frac{{A}_{0}-A}{{A}_{0}}\hspace*{0.28em}\ensuremath{\times{}}\hspace*{0.28em}100\% \end{equation}$

RESULTS AND DISCUSSION

So far as we know, wool is composed of surface-covered cuticle and cortical layers.³² Compared with the latter, the former is much tighter and more difficult to be swollen by formic acid.⁵⁰ To avoid the hindrance of the cuticle layers to the extraction of cortical cells, the former should be removed in advance.

In the current work, the target was achieved by making use of different swelling degrees of cuticle and cortical layers. That is, the wool waste was swollen by formic acid at 100°C for 10 min, and then suffered ultrasonication treatment for 30 min. In the meantime, the poorly swollen cuticle layers burst as a result of the highly swollen cortical layers, so that the broken cuticle layers peeled off subsequently under the impact of ultrasonic wave. As shown in Supporting information Figure S1, after removing the cuticle layers, the wool shows a relatively smooth surface and the scales are no longer visible (Supporting information Figure S1B), which forms striking contrast to the scaly structure of natural wool (Supporting information Figure S1A).

The cortical layer consists of the tightly assembled cortical cells, which contain lots of α-helix crystals and a few amorphous mesenchyme.^32,35,50 Because mesenchyme is loosely compacted in comparison with cortical cells, the disulfide bonds within mesenchyme have to be disrupted in priority when wool waste is oxidized by H₂O₂. In this context, by controlling the concentration of oxidant and oxidation time, the disulfide bonds of mesenchyme can be targetedly destroyed, while the possible damage to the cortical cells is minimized. As a result, cortical cells are allowed to be isolated after the oxidation treatment through the combination of the dissolving effect of formic acid and the ultrasonic aided extraction.

To find out the optimal oxidizing condition and raise the extraction rate, the effects of H₂O₂ concentration and oxidizing time are investigated. The variation in the amount of disulfide bonds of the oxidized wool, which reflect the oxidization degree, are characterized by Raman spectroscopy (Figure 2A and Supporting information Table S1). According to previous works,^{37,46,51–53} it is known that the area ratio of the S-S bond peak (with baseline drawn between 470 and 540/cm) to the internal reference phenylalanine peak (with baseline drawn between 985 and 1020/cm) represents the S-S content. Clearly, when the concentration of H₂O₂ and oxidizing time increase, the content of residual cystine (i.e., disulfide bond) decreases as expected.

[IMAGE OMITTED. SEE PDF]

On the other hand, Figure 2B and Supporting information Table S2 indicate that when the concentration of H₂O₂ is lower or the oxidizing time is not that long, such as 1 wt.% H₂O₂ for 6–18 h and 3 wt.% H₂O₂ for 6–12 h, the amount of the disrupted disulfide bonds and the yield of isolated cortical cells increase as the oxidization proceeds. In the case of higher H₂O₂ concentration or longer oxidation time (3 wt.% H₂O₂ for 18 h or 5 wt.% for 6–18 h), however, the extracting rate of cortical cells decreases as excessive disulfide bonds are destroyed. Particularly, after oxidizing for 12 h by 3 wt.% H₂O₂ solution, about 69.33% of cystine is left and the extraction rate of cortical cells reaches the maximum (i.e., 41.2 wt.%).

Figure 3A-C shows that the extracted cortical cells are spindle-shaped with diameter and length of about 5 and 100 μm, respectively. It is evident that the content of residual cystine of 69.33% is the critical threshold of the oxidation treatment, at which the disulfide bonds of mesenchyme is largely destroyed but those of the cortical cells are mostly remained.

[IMAGE OMITTED. SEE PDF]

Since the extracted cortical cells are planned to act as reinforcement in the regenerated keratin fibers, whose crystalline phases should be characterized as they would greatly affect the mechanical performance of the fibers. Supporting information Figure S2 shows the WAXD patterns of natural wool, the cuticle layers-free wool, oxidized wool, and cortical cells. The natural wool has two characteristic diffraction peaks representing α-helical crystals at 2θ = 9° and β-sheet crystals at 2θ = 20°, respectively. Accordingly, the crystallinities of the materials are calculated and listed in Supporting information Table S3. Compared to natural wool, the relative crystallinity index of the cuticle layers-free wool decreases to 94%, because the formic acid might have slightly degraded the backbones of keratin and break the intermolecular hydrogen bonds and ionic bonds as well. Besides, the relative crystallinity index of the oxidized wool further drops to 86.5% due to cleavage of partial disulfide bonds. As for the cortical cells that are obtained after ultrasonic treatment of the oxidized wool in formic acid, their relative crystallinity index is 64.7%. It means that there are still quite a few crystalline in the cortical cells, which guarantees higher strength of the cortical cells and their capability of reinforcement in the subsequently prepared regenerated keratin fibers.

In addition to strength, toughness of the regenerated keratin fibers should also be considered since pure keratin is fragile. For toughening, PEGDA, a derivative of polyethylene glycol (PEG) that used to play the role of soft segments in polyurethane, is blended with the keratin solution. Its chemical interaction with keratin has been clarified in the Introduction. To assess the toughening effect of PEGDA, the regenerated keratin fibers only containing 0–10 wt.% PEGDA (excluding cortical cells) are prepared. As the mechanical strength of the as-spun fibers is very low, they have to be crosslinked by GA for strengthening prior to the comparison. Figure 4A and Supporting information Table S4 exhibit the tensile behaviors of the GA crosslinked keratin fibers. In the absence of PEGDA, the keratin fibers possess high tensile strength (9.15 ± 1.85 cN/tex) but low elongation at break (5.98 ± 2.09%). When PEGDA is added, tensile strength of the fibers decreases and the elongation at break increases with a rise in the content of PEGDA. At the PEGDA loading of 10 wt.%, for example, tensile strength is 2.85 ± 0.47 cN/tex, while the elongation at break increases to 171.08 ± 18.54%. The data indicate that PEGDA has excellent toughening effect for keratin. Because preparation of high-content keratin fibers is the top priority of this work, 5 wt.% PEGDA is adopted as the representative recipe in the following.

[IMAGE OMITTED. SEE PDF]

It is worth noting that the WAXD patterns of the keratin fibers containing PEGDA without cortical cells (Figure 4B) have no signal at 2θ = 9°. Evidently, all α-helical crystals disappear after the processes of dissolution and wet-spinning, which coincides with the result of our previous work.³⁷

On the basis of the above studies, the as-spun keratin fibers made up of cortical cells (as reinforcement), linear keratin (as adhesive), and PEGDA (as toughening agent) are produced (Figure 1A). The FT-Raman spectra of the as-spun fibers in Supporting information Figure S3 manifest that the absorption at 2570/cm ascribed to sulfhydryl groups (i.e., cysteine) of keratin, which are generated after the decrosslinking and extraction of wool waste, disappears due to the reaction between PEGDA and keratin that has consumed the sulfhydryl groups.

To determine the content of PEGDA that had reacted with keratin, the N contents of the related substances are measured through elemental analysis (Supporting information Table S6). Since the N contents of cortical cells and keratin are approximately identical, the PEGDA content can be calculated from the decrement of N content of the as-spun keratin fibers relative to those of cortical cells and keratin. The results of Supporting information Table S6 indicate that although 5 wt.% PEGDA is added, not all PEGDA are bonded to keratin. Moreover, with increasing the content of cortical cells, the PEGDA content slightly decreases from 4.8 to 3.8 wt.%. In principle, higher content of cortical cells means lower contents of keratin and sulfhydryl groups. As a result, lower content of sulfhydryl groups has to lead to relatively lower reaction efficiency.

From the FT-Raman spectra in Supporting information Figure S3, the content of S-S bonds (relative to natural wool) of cortical cells is estimated to be 92.4% (Table S5) by using the phenylalanine peak as the internal standard, which is much higher than that of the intermediate raw material (i.e. oxidized wool, 69.33%, refer to Supporting information Table S2). It proves that the oxidization before extracting cortical cells mainly leads to fission of the S-S bonds of mesenchyme, while the S-S bonds of cortical cells are mostly reserved. The presence of the peak of S-S bonds on the spectra of the as-spun fibers suggests that the extracted cortical cells have quite complete structure (which is also evidenced by their higher crystallinity [Supporting information Table]) as if they were included in wool. This would help them to function as enhancers in the regenerated fibers.

Figure 4C and Supporting information Table S7 show the tensile properties of GA crosslinked keratin fibers containing 5 wt.% PEGDA and different contents of cortical cells. Compared with the keratin fibers with PEGDA alone (Figure 4A and Supporting information Table S4), the fibers containing both PEGDA and cortical cells have higher tensile strength but lower elongation at break owing to the strengthening effect of cortical cells. When the content of cortical cells is 20 wt.%, the fibers’ tensile strength and elongation at break are 4.78 ± 1.01 cN/tex and 121.80 ± 25.05%, respectively. For the higher content of cortical cells (i.e., 40 wt.%), the fibers’ tensile strength is raised to 8.31±1.02 cN/tex, while the elongation at break decreases to 65.10 ± 12.14%, which are comparable to the tensile properties of natural wool (strength = 8–15 cN/tex, elongation at break = ∼40 %). When the content of cortical cells further increases to 60 wt.%, both tensile strength and elongation at break of the fibers decrease to 7.12 ± 1.02 cN/tex and 20.25 ± 4.58%.

Clearly, the content of cortical cells of 40 wt.% is a turning point. When the content of cortical cells is lower than 40 wt.%, the addition of cortical cells raises tensile strength of the keratin fibers but reduces their elongation at break. The variation trend follows the law of short fibers reinforced polymer composites. When the content of cortical cells exceeds 40 wt.%, there may be too few adhesive (i.e., linear keratin) to effectively transfer stress. Stress concentration would easily appear at the defects in the linear keratin-poor regions, so that the incorporation of cortical cells results in decrease of both tensile strength and elongation at break of the fibers.

The above analysis receives support from the surface morphologies of the fibers (Supporting information Figures S4, S5A and B). In case no cortical cells are present, the fibers surfaces are rather smooth (Supporting information Figure S4A and B). Upon the addition of cortical cells (e.g., at the content of 20 wt.%), their contours start to be discernible (Supporting information Figure S4C and D). When the content of cortical cells is raised to 40 wt.%, the outlines of cortical cells can be clearly seen from the fibers surfaces, but the fibers still maintain the compact structure without obvious flaws (Supporting information Figure S5A and B). The observation results well agree with the effectiveness of cortical cells serving as reinforcements (Figure 4C). As the content of cortical cells reaches 60 wt.%, the fibers surfaces look rather coarse. Not only apparent defects but also de-bonded cortical cells are perceived (Supporting information Figure S4E and F). In fact, the spinning process becomes quite difficult at the high cortical cells loading. Therefore, the mechanical properties of the keratin fibers have to be poor under the circumstances (Figure 4C). The cross-sectional morphologies of the fibers (Supporting information Figure S3D-K) further reveal their microstructural compactness as a function of cortical cells content from another angle. Especially, for the fibers with 60 wt.% cortical cells, they look looser than those containing smaller amount of cortical cells.

On the whole, the SEM images of Supporting information Figure S4C-F and S5A-B demonstrate that the cortical cells are aligned along the fiber axis owing to the spinning process. This is critical for the cortical cells to jointly take effect in the axial direction.

The WAXD patterns in Figure 5A reveal the crystalline structures of the GA crosslinked keratin fibers containing 5 wt.% PEGDA and different contents of cortical cells. Compared with natural wool and cortical cells, the peaks of β-sheet crystals (2θ = 20°) of the fibers are remarkably enhanced, while the peaks of α-helical crystals (2θ = 9°) of the fibers still exist due to the appearance of cortical cells, despite they are significantly weakened. A careful survey of Figure 5A suggests that the intensities of the peaks at 2θ = 9° increase as the content of cortical cells increases. Accordingly, the crystallinity index (Supporting information Table S8) is changed from 0.094 to 0.283 with a rise in the content of cortical cells from 20 to 60 wt.%. It implies that the regenerated fibers with cortical cells arranged along the fiber axis have acquired wool-like structure, which contains α-helical crystals of the intermediate filaments as reinforcement.

[IMAGE OMITTED. SEE PDF]

Returning to Figure 4C and Supporting information Table S7, although the GA crosslinked keratin fibers containing 5 wt.% PEGDA and 40 wt.% cortical cells have decent elongation at break, their tensile strength seems to be slightly insufficient in comparison with natural wool. To further increase the fibers’ strength, therefore, MDI is employed as the second crosslinking agent, which is able to react with the hydroxyl or amino groups of the fibers (Figure 1F). Accordingly, continuous keratin fibers crosslinked by GA/MDI (Figure 1H) are fabricated from discontinuous wool waste (Figure 1G). The enlarged views of the fibers (Supporting information Figure S5C and D) indicate that the diameter is about 50 μm, which is close to the values of the GA crosslinked fibers (Supporting information Figure S5A and B) and natural wool. Besides, their surfaces are also rather rough, similar to the GA crosslinked versions, and the outlines of cortical cells can be seen (Supporting information Figure S5). Nevertheless, the fibers have tight structure without obvious defects as reflected by the cross-sectional view (Supporting information Figure S6), and the cortical cells are still arranged along the fiber axis.

Consequently, tensile strength of the dually crosslinked fibers increases to 10.03 ± 1.21 cN/tex, which is higher than those of wool waste (6.06 ± 2.22 cN/tex) and GA crosslinked keratin fibers (8.31 ± 1.02 cN/tex). Meanwhile, the elongation at break decreases to 42.54 ± 8.54%, which is lower than that of GA crosslinked keratin fibers (65.10 ± 12.14), but close to that of wool waste (41.3 ± 10.8%) (Figure 4D and Supporting information Table S9). Such a variation must be correlated to the increase of crosslinking density. As shown in Supporting information Table S10, the GA crosslinked keratin fibers have a higher equilibrium swelling degree of 2.17 ± 0.27 than that of natural wool (1.60 ± 0.004), and the fibers crosslinked by both GA and MDI possess a lower equilibrium swelling degree of 1.58 ± 0.11. It can thus be concluded that by taking advantage of the treatment with the two kinds of crosslinking agent, the regenerated keratin fibers coupled with a few PEGDA (5 wt.%) and 40 wt.% cortical cells have acquired mechanical properties resembling those of natural wool, showing greater potential for application as textile material than the keratin fibers crosslinked only by GA.

Since the crosslinking treatment has shown its ability in producing high-performance keratin fibers, the amounts of the crosslinking agents that had reacted with keratin fibers should be known by weighing the samples before and after crosslinking, besides quantification of its effect on the fibers’ mechanical properties. As listed in Supporting information Table S11, 1.69 ± 0.29 wt.% GA and 1.02 ± 0.14 wt.% MDI have been successively attached to the recycled fibers. Considering that the actual content of the bonded PEGDA is 4.0 wt.% in this case (Supporting information Table S6), the total content of keratin and cortical cells in the regenerated fibers is estimated to be 93.3 wt.%. Compared with the works of other groups,^{3,13,14,28–30} the regenerated fibers obtained in this work possess not only higher content keratin and excellent tensile strength, but also similar elongation at break like natural wool (Table 1).

TABLE 1 Comparison of our keratin fibers with those reported by other groups

References	Keratin type	Content of keratin (wt.%)	Tensile strength (cN/tex)	Elongation at break (%)
Yang et al.3	Wool	100	7.77	10.9
Yang et al.28	Feather	100	5.77	5.0
Yang et al.30	Feather	100 (except crosslinking agent)	12.3	14
Fan et al.29	Feather	100	2	3
Katoh et al.13	Wool	45	7.7	12.6
Liu et al.14	Wool	25	4.6	38.4
Our previous work37	Wool	86	9.21	39.20
This work	Wool	93.3	10.03	42.54

More details of the variations in the chemical structures aroused by the crosslinking are disclosed by Raman spectra (Supporting information Figure S7). It is seen that the characteristic peaks of C = N bonds of Schiff base, C = O groups of carbamates, and benzene rings of MDI (1550–1750/cm, i.e. the amide I bands) are formed after the crosslinking reaction between the amino and hydroxyl groups of wool keratin and aldehyde groups of GA/isocyanates of MDI. Referring to the absorption of phenylalanine at 1003/cm as the internal standard,^46,51–53 the related reaction degrees can be known (Supporting information Table S12). Owing to the generation of C = N, for example, the peak area ratio of the amide I band of the GA crosslinked fibers is raised from 22.99 of the as-spun fibers to 32.84, and it is further increased to 43.32 after the crosslinking with MDI. Meantime, the peak area ratio of the skeletal vibration of the benzene ring of MDI at 1510/cm increases from 0 of the as-spun fibers and GA crosslinked fibers to 0.68.

The infrared spectra in Supporting information Figure S8 reveal the information of the crosslinking at a different angle. The peaks at 1660 and 1541/cm are attributed to the C = O stretching and N-H bending in the amide bonds of keratin, respectively. After crosslinking by GA, a few amino groups at 1541/cm are consumed, giving rise to the peak of C = N overlapping that of C = O on the spectrum of the fibers. Accordingly, the intensity ratio of the peak at 1541/cm to that at 1660/cm decreases from 0.91 to 0.83 (Supporting information Table S13). When the second crosslinking reaction takes place with the aid of MDI, the ratio further decreases to 0.80 (Supporting information Table S13) owing to the generation of C = O.

Based on these studies of chemical structures, we can discuss the crystalline feature of the regenerated keratin fibers better. As can be seen from Figure 5B, the peak at 2θ = 9° assigned to α-helical crystals of the intermediate filaments of wool appears again on the WAXD spectrum of the regenerated fibers containing 5 wt.% PEGDA and 40 wt.% cortical cells, forming contrast to the fibers without cortical cells (Figure 4B). The relative crystallinity index of the regenerated fibers crosslinked by GA/MDI is 25.3% (Supporting information Table S14), which is slightly lower than the value of the fibers before crosslinking because a few α-helical crystals are damaged by the organic reagents during the crosslinking. Nevertheless, the crystallinity indexes of the regenerated fibers still roughly obey the principle of linear superposition. It means that the addition of cortical cells has introduced α-helical crystals to the regenerated fibers, which are responsible for the strengthening effect.

The orientations of crystalline phases in the keratin fibers are also determined by WAXD (Supporting information Figure S9). Since the keratin fibers exhibit a more pronounced peak at 20° originating from β-sheet crystals (Figure 5B), while natural wool shows a more pronounced peak at 9° attributing to α-helix crystals, the full widths at half maximums of the WAXD curves can be used to calculate the orientation degrees (Supporting information Table S15) by fixing 2θ at 9° and 20° and rotating the sample for 180°. The results suggest that the crystals in keratin fibers are slightly more oriented than the crystals in natural wool. The oriented crystallines help to bring the strengthening effect of cortical cells into play in the axial direction, which well agrees with the SEM observation of the fibers surface morphologies (Supporting information Figure S5C and D).

In addition to the crystalline structure, the secondary structure of keratin chains is also correlated to the mechanical properties of the regenerated fibers. Accordingly, the ¹³C NMR spectra of the fibers (Figure 5C) are collected to investigate the protein conformations including α-helical chains, β-folding chains, and random coils.^26,27,47 The peaks at 54 and 40 ppm are attributed to α-carbon and β-carbon of amino acid, respectively. As for the peak at around 173 ppm, it represents the carbonyl groups and can be decomposed into two subpeaks (Figure 5D), one at 172 ppm associated with β-folding and random coils, and the other at 176 ppm attributed to the α-helix chains. The curve fitting results (Supporting information Table S16) indicate that the content of α-helical chains of natural wool is up to 88.4%, which is in accordance with previous reports.^26,37,47 The conformations of the extracted linear keratin are mainly β-folding chains and random coils, and the content of α-helical chains is only 4.9%. With respect to the extracted cortical cells, its content of α-helical chains is 72.6%, lower than that of natural wool, because some α-helical chains turn into β-folding chains and random coils in the course of oxidation, swelling by formic acid, and ultrasonic treatment. When comes to the as-spun fibers without cortical cells, the addition of PEGDA increases the content of α-helical chains, probably its chemical interaction (i.e., hydrogen bonds) with keratin replaces the interaction among keratin molecules, and hence, the transformation of keratin conformation from β-folding to α-helical chains is favored. Upon being stretched, the α-helical chains would be straightened, dissipating the input mechanical energy. Besides, PEGDA is a PEG derivative, while PEG usually acts as soft segments in polyurethane to improve its toughness. In this context, both PEGDA and the α-helical chains would contribute to the improvement of toughness of the as-spun regenerated fibers. When cortical cells are added, the α-helical chains content of the as-spun fibers increases to 39.5%, and then slightly decreases to 36.7% after crosslinking by GA/MDI. These α-helical chains and β-folding chains of the linear keratin must be conducive to the increase of toughness and strength of the fibers, respectively.

For purposes of finding opportunities for practical application, a few nonstructural properties of the regenerated keratin fibers are characterized (Table 2). The UV spectra of the dye liquor before dying and the residual dye liquors after dyeing are given in Supporting information Figure S10, which can be used for calculating the dye uptake. Because of consumption of amino groups after crosslinking by GA, the dyeing performance of the regenerated fibers is marginally poorer than that of natural wool (Table 2). As for the hygroscopic capacity expressed by standard moisture regain, it is also lower than natural wool but close to cotton,⁵⁴ due to the reduction of hydrophilic groups during the crosslinking reaction with GA and MDI. On the other hand, the electrical conductivity of the regenerated keratin fibers is higher than wool, representing improved antistatic property. This is because of the tightly stacked microstructure of the former, which contrasts to the latter that contains plenty of inherent cavities.

TABLE 2 Dyeing performance, standard moisture regain and volume resistivity of the regenerated fibers and natural wool

Sample	Dye-uptake (%)	Standard moisture regain (wt%)	Volume resistivity (MΩ cm)
Regenerated fibers*	77.6	8.70 ± 0.38	287.3 ± 15.2
Natural wool	83.2	12.36 ± 0.62	310.4 ± 10.3

The TGA curves of the related substances are plotted in Supporting information Figure S11 to investigate their thermal stability. Compared with natural wool and cortical cells, the regenerated fibers have slightly lower onset pyrolysis temperature because of less crystalline phases. Moreover, the weight loss at around 100°C originates from water desorption. Since the hydrophilic amino groups and hydroxyl groups are consumed after crosslinking, the hygroscopic capacity of the fibers has to be weakened. The result coincides with the measurements of moisture regain in Table 2.

CONCLUSIONS

The present work developed a novel method for recycling wool waste and preparing continuous regenerated keratin fibers. By optimizing the oxidization conditions, spindle-shaped cortical cells were successfully obtained from wool waste at the extraction ratio of 41.2 wt.%. Meanwhile, linear keratin was produced by dissolving wool waste with a green solution system of DTT/SDS. Then, continuous regenerated keratin fibers were fabricated via wet-spinning, in which the cortical cells served as reinforcement, linear keratin as adhesive, PEGDA as toughener, and GA and MDI as crosslinking agents, respectively. The total content of cortical cells and keratin was up to 93.3 wt.%, which is higher than the result of our earlier attempt in this aspect based on another technical route.³⁷ As a result, the content of the synthetic chemicals in the fibers were reduced to a rather low level.

The cortical cells introduced α-helical crystals of the intermediate filaments into the regenerated fibers, which provided the fibers with wool-like structure and improved the tensile strength. Through the synergistic effects of different components, the regenerated keratin fibers possessed properties comparable to those of natural wool including mechanical properties, thermal stability, dyeing performance, antistatic property, and moisture absorption. The present work has bionic significance. It may be particularly attractive for recycling animal hairs with similar compositions and structures^.55–58

ACKNOWLEDGMENTS

The valuable instruction of Professor Weilin Xu from Wu Han Textile University, China is highly appreciated. The authors thank the support of the National Natural Science Foundation of China (Grants: 52033011, 51773229, 51873235 and 51973237). Natural Science Foundation of Guangdong Province (Grants: 2019B1515120038, 2021A1515010417 and 2020A1515011276), Science and Technology Planning Project of Guangdong Province (Grant: 2020B010179001), and Industry-University-Research Collaboration Project of Zhuhai City (Grant: ZH22017001200004PWC).

CONFLICT OF INTEREST

The authors declare no conflict of interest.

AUTHOR CONTRIBUTIONS

The manuscript was written through contributions of all authors. All authors have given approval to the final version of the manuscript.

DATA AVAILABILITY STATEMENT

All data supporting the findings of this study are available within the article and the Supplementary Information file. All data are available on request from the corresponding authors.

References

Navone L, Moffitt K, Hansen K, Blinco J, Payne A, Speight R. Closing the textile loop: enzymatic fibre separation and recycling of wool/polyester fabric blends. Waste Manage. 2020;102:149‐160.

Hassan MA, Abol‐Fotouh D, Omer AM, Tamer TM, Abbas E. Comprehensive insights into microbial keratinases and their implication in various biotechnological and industrial sectors: a review. Int J Biol Macromol. 2020;154:567‐583.

Xu H, Ma Z, Yang Y. Dissolution and regeneration of wool via controlled disintegration and disentanglement of highly crosslinked keratin. J Mater Sci. 2014;49(21):7513‐7521.

Sandin G, Peters GM. Environmental impact of textile reuse and recycling: a review. J Cleaner Prod. 2018;184:353‐365.

Barba C, Méndez S, Roddick‐Lanzilotta A, Kelly R, Parra JL, Coderch L. Cosmetic effectiveness of topically applied hydrolysed keratin peptides and lipids derived from wool. Skin Res Technol. 2008;14(2):243‐248.

Nikiforova T, Kozlov V, Islyaikin M. Sorption of d‐metal cations by keratin from aqueous solutions. J Environ Chem Eng. 2019;7(5): [eLocator: 103417].

Giuri D, Barbalinardo M, Sotgiu G, et al. Nano‐hybrid electrospun non‐woven mats made of wool keratin and hydrotalcites as potential bio‐active wound dressings. Nanoscale. 2019;11(13):6422‐6430.

Agarwal V, Panicker AG, Indrakumar S, Chatterjee K. Comparative study of keratin extraction from human hair. Int J Biol Macromol. 2019;133:382‐390.

Saul JM, Ellenburg MD, de Guzman RC, Dyke MV. Keratin hydrogels support the sustained release of bioactive ciprofloxacin. J Biomed Mater Res, Part A. 2011;98A(4):544‐553.

Shavandi A, Silva TH, Bekhit AA, Bekhit AEA. Keratin: dissolution, extraction and biomedical application. Biomater Sci. 2017;5(9):1699‐1735.

Aluigi A, Vineis C, Ceria A, Tonin C. Composite biomaterials from fibre wastes: characterization of wool‐cellulose acetate blends. Composites, Part A. 2008;39:126‐132.

Xie H, Li S, Zhang S. Ionic liquids as novel solvents for the dissolution and blending of wool keratin fibers. Green Chem. 2005;7(8):606.

Katoh K, Shibayama M, Tanabe T, Yamauchi K. Preparation and properties of keratin‐poly(vinyl alcohol) blend fiber. J Appl Polym Sci. 2004;91(2):756‐762.

Liu R, Li L, Liu S, et al. Structure and properties of wool keratin/poly (vinyl alcohol) blended fiber. Adv Polym Tech. 2018;37(8):2756‐2762.

He C, Hu Y, Wang Y, et al. Complete waste recycling strategies for improving the accessibility of rice protein films. Green Chem. 2020;22(2):490‐503.

Gambucci M, Aluigi A, Seri M, et al. Effect of chemically engineered Au/Ag nanorods on the optical and mechanical properties of keratin based films. Front Chem. 2020;8:158.

Song K, Qian X, Zhu X, Li X, Hong X. Fabrication of mechanical robust keratin film by mesoscopic molecular network reconstruction and its performance for dye removal. J Colloid Interf Sci. 2020;579:28‐36.

Wu S, Chen X, Li T, et al. Improving the performance of feather keratin/polyvinyl alcohol/tris(hydroxymethyl)aminomethane nanocomposite films by incorporating graphene oxide or graphene. Nanomaterials. 2020;10(2):327.

Valkov A, Zinigrad M, Sobolev A, Nisnevitch M. Keratin biomembranes as a model for studying onychomycosis. Int J Mol Sci. 2020;21(10):3512.

Mi X, Xu H, Yang Y. Submicron amino acid particles reinforced 100% keratin biomedical films with enhanced wet properties via interfacial strengthening. Colloids Surf B. 2019;177:33‐40.

Cera L, Gonzalez GM, Liu Q, et al. A bioinspired and hierarchically structured shape‐memory material. Nat Mater. 2020;20:242‐249.

Wang K, Li R, Ma JH, Jian YK, Che JN. Extracting keratin from wool by using L‐cysteine. Green Chem. 2016;18(2):476‐481.

Shavandi A, Bekhit AEA, Carne A, Bekhit A. Evaluation of keratin extraction from wool by chemical methods for bio‐polymer application. J Bioact Compat Pol. 2017;32(2):163‐177.

Feroz S, Muhammad N, Ratnayake J, Dias G. Keratin ‐ Based materials for biomedical applications. Bioact Mater. 2020;5(3):496‐509.

Rajabinejad H, Zoccola M, Patrucco A, Montarsolo A, Rovero G, Tonin C. Physicochemical properties of keratin extracted from wool by various methods. Text Res J. 2017;88:2415‐2424.

Zhang Z, Nie Y, Zhang Q, et al. Quantitative change in disulfide bonds and microstructure variation of regenerated wool keratin from various ionic liquids. ACS Sustain Chem Eng. 2017;5(3):2614‐2622.

Wang L, Nie Y, Zhang X, Zeng S, Zhang S, Zheng S. Synergistic effects of cosolvents on the dissolution of wool keratin using ionic liquids. Chem Eng Technol. 2016;39(5):979‐986.

Xu H, Yang Y. Controlled de‐cross‐linking and disentanglement of feather keratin for fiber preparation via a novel process. ACS Sustain Chem Eng. 2014;2(6):1404‐1410.

Fan X, Value‐added products from chicken feather fiber and protein. PHD Thesis. Auburn University, 2008.

Mi X, Li W, Xu H, Mu B, Chang Y, Yang Y. Transferring feather wastes to ductile keratin filaments towards a sustainable poultry industry. Waste Manage. 2020;115:65‐73.

Aradi E, Naidoo SR, Billing DG, Wamwangi D, Motochi I, Derry TE. Ion beam modification of the structure and properties of hexagonal boron nitride: an infrared and X‐ray diffraction study. Nucl Instrum Meth B. 2014;331:140‐143.

Höcker H. Fibre morphology. In: Simpson WS, Crawshaw GH, eds. Wool‐science and technology (pp. 60—79). Woodhead Publishing Ltd; 2002.

Fan J, Yu W. Biomaterials from wool wastes: characterization of cortical cells/chitosan composite. Waste Manage Res. 2010;28(1):44‐50.

Liu Y, Yin R, Yu W. The bio‐inspired study of homogeneous composite materials. J Compos Mater. 2011;45(1):113‐125.

Patrucco A, Aluigi A, Vineis C, Tonin C. Bio‐composite keratin films from wool fibrillation. J Biobased Mater Bio. 2011;5(1):124‐131.

Hearle JWS. Physical properties of wool. In: Simpson WS, Crawshaw GH, eds. Wool‐science and technology (pp. 80–129). Woodhead Publishing Ltd; 2002.

Cao G, Rong MZ, Zhang MQ. Continuous high‐content keratin fibers with balanced properties derived from wool waste. ACS Sustain Chem Eng. 2020;8(49):18148‐18156.

Wu Y, Simpson MC, Jin J. Fast hydrolytically degradable 3D printed object based on aliphatic polycarbonate thiol‐yne photoresins. Macromol Chem Phys. 2021;222(6): [eLocator: 2000435].

Pineiro‐Garcia A, Vega‐Diaz SM, Tristan F, Meneses‐Rodriguez D, Labrada‐Delgado GJ, Semetey V. New insights in the chemical functionalization of graphene oxide by thiol‐ene Michael addition reaction. Flatchem. 2021;26(3): [eLocator: 100230].

Love D, Fairbanks B, Bowman C. Reaction environment effect on the kinetics of radical thiol‐ene polymerizations in the presence of amines and thiolate anions. ACS Macro Lett. 2020;9(2):174‐179.

Gao H, Sun Y, Wang M, et al. Self‐healable and reprocessable acrylate‐based elastomers with exchangeable disulfide crosslinks by thiol‐ene click chemistry. Polymer. 2021;212: [eLocator: 123132].

Antony R, Arun T, Manickam STD. A review on applications of chitosan‐based Schiff bases. Int J Biol Macromol. 2019;129:615‐633.

Wang C, Zheng Y, Sun Y, Fan J, Qin Q, Zhao Z. A novel biodegradable polyurethane based on poly(3‐hydroxybutyrate‐co‐3‐hydroxyvalerate) and poly(ethylene glycol) as promising biomaterials with the improvement of mechanical properties and hemocompatibility. Polym Chem. 2016;7(39):6120‐6132.

Feula A, Tang X, Giannakopoulos I, et al. An adhesive elastomeric supramolecular polyurethane healable at body temperature. Chem Sci. 2016;7(7):4291‐4300.

Long J, Cui C, Wang L, Xu H, Yu Z, Bi X. Effect of treatment pressure on wool fiber in supercritical carbon dioxide fluid. J Clean Prod. 2013;43:52‐58.

Song YX, Rong MZ, Zhang MQ. Improvement of multiple‐responsive shape memory effects of wool through increasing the content of disulfide bonds. Polymer. 2020;188: [eLocator: 122130].

Jiang Z, Yuan J, Wang P, et al. Dissolution and regeneration of wool keratin in the deep eutectic solvent of choline chloride‐urea. Int J Biol Macromol. 2018;119:423‐430.

Zhao Q, Feng H, Wang L. Dyeing properties and color fastness of cellulase‐treated flax fabric with extractives from chestnut shell. J Clean Prod. 2014;80:197‐203.

Liu M, Chen J, Zhang J, Cao Y. The optimization of reactive dyeing process of wool. Wool Text J. 2019;47:37‐40.

Fan J, Yu W. Basic parameter study for the separation of wool fibre components by ultrasonic irradiation in formic acid. Res J Text Apparel. 2009;13(2):69‐74.

Kuzuhara A. Chemical modification of keratin fibers using 2‐iminothiorane hydrochloride. J Appl Polym Sci. 2003;90(13):3646‐3651.

Kuzuhara A. Protein structural changes in keratin fibers induced by chemical modification using 2‐iminothiolane hydrochloride: a Raman spectroscopic investigation. Biopolymers. 2005;79(4):173‐184.

Kuzuhara A, Hori T. Reducing wrinkle formation in wool with 2‐iminothiorane hydrochloride. Text Res J. 2002;72(4):285‐289.

Zhang J, France P, Radomyselskiy A, Datta S, Zhao J, van Ooij W. Hydrophobic cotton fabric coated by a thin nanoparticulate plasma film. J Appl Polym Sci. 2003;88(6):1473‐1481.

Yu S, Xiang H, Zhou J, Qiu T, Hu Z, Zhu M. Typical polymer fiber materials: an overview and outlook. Acta Polym Sin. 2020;51(1):39‐54.

Yang G, Gu K, Shao Z. The investigation from animal silks to silk protein‐based materials. Acta Polym Sin. 2021;52:16‐28.

Zhong Y, Cai J, Zhang L. A review of chitin solvents and their dissolution mechanisms. Chinese J Polym Sci. 2020;38:1047‐1060.

Chen T, Zhong G, Zhang Y, Zhao L, Qiu Y. Bio‐based and biodegradable electrospun fibers composed of poly(l‐lactide) and polyamide 4. Chinese J Polym Sci. 2020;38:53‐62.