Ultrasonic-assisted refinement of domesticated-wild silk protein composite nanofibers: enhancing miscibility, uniformity, and functionality via ionic liquid processing

Abstract

Silk fibroin is highly regarded for its exceptional biocompatibility, degradability, and mechanical properties, making it a valuable material in the field of tissue engineering. Ultrasound technology, recognized as a safe and efficient physical method, enables precise manipulation of material microstructures and macroscopic properties, which is essential for the development of innovative high-performance biomaterials. This study aims to enhance the solution miscibility, fiber uniformity, and properties of silk-based protein nanofiber materials by employing a silk-silk composite approach. The method involved air-spinning of Tussah silk fibroin (TSF) and Bombyx mori silk fibroin (BSF) blends through ionic liquid dissolution, combined with ultrasound-assisted processing. Comprehensive characterization, including SEM, FTIR, XRD, ¹³C NMR, DSC, TGA, AFM, WCA, revealed that the original TSF/BSF composite exhibited weak hydrogen bonding interactions, resulting in uneven protein fibers. Moderate ultrasound treatment facilitated the formation of uniform fibers and their interlacing, significantly enhancing the interactions between TSF and BSF. This process promoted the efficient miscibility of TSF with BSF, thereby mitigating the occurrence of microphase separation. It led to increased β-sheet crystalline content, improved thermal and mechanical properties, and enhanced hydrophilicity, biocompatibility, and biodegradation rates. Therefore, integrating protein composites with ultrasound processing produces uniform nanofiber biomaterials with superior structural and biological properties, opening up new perspectives for their application in biomedicine.

1. Introduction

Silk, a natural protein fiber produced by both wild and domestic silkworms, has been widely used in the textile industry for centuries [1,2]. It primarily consists of silk fibroin (70–80 wt%) and sericin (20–25 wt%), with silk fibroin typically extracted through heat or alkaline treatments [3,4]. Silks from silkworm cocoons have various types, such as Antheraea mylitta (Tussah), Philosamia ricini (Eri), and Bombyx mori mulberry (Mori) [4,5]. Known for its chemical stability, mechanical strength, environmental friendliness, biocompatibility and biodegradability,and hypoallergenic properties, silk fibroin has found extensive application in tissue engineering and biomedicine [2,3,5]. Particularly, wild tussah silk fibroin (TSF) and Bombyx mori silk fibroin (BSF) exhibit significant differences in structure and function. The crystalline regions of TSF, rich in polar groups, primarily consist of alanine, glycine, serine, and hydrophilic aspartic acid [6,7]. In contrast, the crystalline regions of BSF predominantly consist of repeating glycine-alanine sequences, and the molecule is composed of heavy chains (H-chains), light chains (L-chains), and P25 protein in a ratio of 6:6:1 [8,9], which partially endows BSF with superior extensibility and film processing properties [10,11]. However, TSF lacks the L-chain and P25 protein. The hydrophobic structure of BSF is formed by the Gly-X (X representing Ala, Ser, Thr, Val) repeating sequences, which can form antiparallel β-sheets [6]. TSF displays a higher alanine/glycine ratio and numerous poly-alanine segments to form β-sheets [12,13]. Additionally, the mechanical properties of regenerated TSF materials are poor after dissolution, limiting their application in tissue engineering [14]. In terms of biocompatibility, the Arg-Gly-Asp (RGD) sequence in TSF confers greater hydrophilicity and cellular adhesion [7,14]. These findings indicate that the structural differences between BSF and TSF significantly influence the overall performance of silk fibroin-based regenerative materials. Therefore, blending regenerated natural TSF and BSF is an important approach, which leverages the characteristics of each component to enhance the properties of silk fibrous materials. Yang et al. [8] successfully prepared a TSF/BSF composite membrane using CaCl₂/formic acid as the cosolvent and water as the post-treatment solvent. Compared to pure BSF and TSF membranes, this composite membrane exhibited higher thermal stability and controlled hydrophilicity. When the TSF:BSF (w/w) ratio was 3:1, the composite membrane demonstrated excellent mechanical properties and biocompatibility. Our results confirmed these findings: Various mass ratios (e.g., 1:1, 1:3) of composite solutions were used for air spinning; however, these attempts failed to produce a stable material. Subsequently, only the TSF:BSF ratio of 3:1 in the composite air spinning process successfully led to the formation of a fibrous membrane. This ratio preserved the inherent advantages of TSF while effectively incorporating the superior properties of BSF. Therefore, these two representative and widely used types of silk are selected for this research, which can provide a basis for further studies and applications of other types of silk in the future.

Ionic liquids (ILs) are a unique class of liquid salts consisting of organic cations and inorganic or organic anions. Their distinctive physical and chemical properties enable the efficient dissolution of biomacromolecules such as silk, cellulose, and lignin [[15], [16], [17]], facilitating biomass utilization and conversion. In addition, these properties allow for the recovery and reuse of the solvent from the used material, addressing resource and environmental challenges compared to traditional preparation methods using acids [6]. For instance, Chen et al. [18] developed a protonic IL [DBNH][Mea], capable of dissolving silk protein and cellulose under mild conditions. Using wet spinning, they successfully produced highly compatible silk/cellulose composite fibers, promoting sustainable biomaterials. Similarly, Zhang et al. [19] used ILs to separate and purify silkworm pupal protein (SPP), achieving a 62.6 % extraction yield with less than 0.5 % fat content, while preserving SPP’s original structure. This method enhances SPP quality and applications while providing a sustainable, eco-friendly alternative to conventional processes. Specifically, 1-Ethyl-3-methylimidazolium acetate (EMIMAc) is an advanced IL renowned for its superior ability to dissolve proteins. Unlike other ionic liquids, EMIMAc enables the dissolution of silk fibroin (SF) more gently, thereby minimizing structural damage and preserving the native conformation and biological activity of SF [20,21].

Phase separation is a detrimental phenomenon in polymer blends, leading to inhomogeneous fiber materials with inconsistent diameters. This structural separation occurs due to thermodynamic incompatibility among different components within a material [22,23]. While chemical bonding between different monomers prevents macroscopic phase transitions, phase-separated regions still form at the nanoscale to microscale range [24]. Such phase separation in composite materials can be highly undesirable, as it disrupts structural uniformity and compromises material properties [25,26], leading to irregular fiber formation and variations in material characteristics, ultimately reducing the reliability and effectiveness of the final product.

Ultrasonic processing is a fast, efficient, and precise material treatment technique [27,28]. It uses high-frequency sound waves (above 20 kHz) to generate cavitation, releasing intense shockwaves, heat, and pressure. This process enhances solute dispersion, reduces phase separation in composites, and induces polymer chain scission or reorganization, improving material properties [27,29,30]. This simple and effective physical process for preparing nanofibers benefits the biological environment [29,30]. Combined with solution spray spinning, an advanced fiber manufacturing method, it enables the production of micron- or nanoscale fibers by spraying a polymer solution into a high-speed air stream [31,32]. Compared to traditional electrospinning and melt spinning, solution-jet spinning offers lower energy consumption, improved safety, and superior environmental and cost performance[33].

In our previous study [34], we successfully integrated ultrasonic processing with solution spinning to fabricate TSF nanofiber membranes with uniform morphology and excellent structural properties. This approach not only improved fiber uniformity and stability but also enhanced the mechanical strength and biocompatibility of the membranes, contributing to the development of high-performance biomaterials. This study marks the first investigation into the jet-spun fabrication of protein–protein composite fiber membranes from ionic liquids, using TSF/BSF as a model system. It also compares the structures and properties of these composite membranes with those of two individual protein membranes, highlighting the unique advantages of the composites. Additionally, the study explores the effects of ultrasonic treatment on improving the performance of composite fiber membranes and its impact on the phase separation behavior of silk. The surface morphology of the recycled SF fiber membranes was examined using scanning electron microscopy (SEM). Molecular and crystal structures were analyzed through Fourier transform infrared spectroscopy (FTIR), X-ray powder diffraction (XRD), and 13C nuclear magnetic resonance spectroscopy (¹³C NMR). Thermal, mechanical, and hydrophilic properties were evaluated using differential scanning calorimetry (DSC), thermogravimetric analysis (TGA), atomic force microscopy (AFM), and water contact angle (WCA) measurements. Finally, the biological properties of the material were assessed through cell viability and enzymatic degradation assays.

2. Experiments

2.1. Raw materials

The raw wild Tussah cocoons and Bombyx mori cocoons were supplied by July Trading Co., Ltd. (Liaoning, China). Sodium bicarbonate (NaHCO₃, CAS: 144-55-88, Purity: 99.5 %, Analytical-grade) was procured from Sinopharm Chemical Reagent Co., Ltd. (China). Additionally, the ionic liquid 1-Ethyl-3-methylimidazolium acetate (EMIMAc, CAS: 143314-17-4, Purity: 98 %, Analytical-grade) was provided by Shanghai Chengjie Chemical Co., Ltd. (China). Deionized water (CAS: 7732-18-5, Laboratory Grade II Pure Water) was obtained from the School of Chemistry and Materials Science at Nanjing Normal University (China).

2.2. Sample preparation

Firstly, using NaHCO₃ to degum silkworm cocoons, yielding dried regenerated silk fibroin proteins. These proteins were then dissolved in EMIMAc ionic liquid at 125 °C to form 8.00 (w/w) solutions of Tussah silk fibroin (TSF), Bombyx mori silk fibroin (BSF), and their 3:1 (w/w) blend. The mixed silk fibroin solution was ultrasonically treated using an ultrasonic cell crusher (WM-1000 T, Shanghai Weimi Technology Co., Ltd., China) for 30 min at power levels of 0 W, 150 W, 300 W, 450 W, and 600 W, while maintaining the temperature below 40 ± 0.5°C with a circulating ice-water bath. After treatment, the solution was processed using a jet spinning equipment (DQE750-24L, Jiangsu Dongcheng Electromechanical Co., Ltd., China) under an air pressure of 0.9 ± 0.01 MPa, a liquid inlet rate of 30 ± 0.3 mL/h, and a nozzle-to-collector distance of 0.5 ± 0.05 m. Finally, the regenerated SF fiber membrane samples were washed in deionized water and dried (Fig. 1). The samples were named based on sonication power: TSF-0 and BSF-0 for the untreated single-protein samples, and TB-0, TB-150, TB-300, TB-450, and TB-600 for the composite protein samples treated at different power levels.

Fig. 1.

Process flow chart of preparing TSF/BSF composite fiber membrane based on EMIMAc ultrasonic spray spinning.

2.3. Materials characterization

2.3.1. Scanning electron microscopy

The surface morphology of the regenerated SF fiber membrane was observed using a Regulus 8100 cold field emission scanning electron microscope (Hitachi, Japan). To enhance the electrical conductivity of the samples, they were uniformly coated with Pt/Pd for 20 s using a JFC-1600 sputter coater (JEOL, Japan). After coating, the samples were placed in the SEM chamber, evacuated to a vacuum, and subsequently examined. SEM images were processed using ImageJ 1.53e software to determine the single fiber diameter (d_i) and single fiber surface area (A_i). The formulas for calculating the average fiber diameter (D) and average surface area (A) are as follows:

$$D=\frac{{\sum }_{i=1}^i{d}_i}{i}$$ (1)

$$A=\frac{{\sum }_{i=1}^i{A}_i}{n}$$ (2)

where i represents the number of fibers with diameter d_i, and n denotes the number of analysis iterations for the sample.

2.3.2. Fourier transform infrared spectroscopy

The structural characterization of TSF/BSF composite fiber membranes prepared at various ultrasonic power levels was performed using a NEXUS-670 Fourier Transform Infrared Spectrometer (Thermo Nicolet, USA). The spectral range was set from 4000 cm⁻¹ to 400 cm⁻¹, with 64 continuous scans at a resolution of 4 cm⁻¹. The acquired infrared spectra underwent baseline correction, atmospheric correction, and normalization. The SF amide I region was then fitted with a Gaussian function using Origin software to quantitatively analyze its secondary structure. The fitting process was repeated five times, maintaining a Chi² factor below 10⁻⁴.

2.3.3. X-ray diffraction analysis

The crystal structures of TSF/BSF composite fiber membranes was tested using a D/MAX-2500PC X-ray diffractometer (Rigaku, Japan). The sample was placed on a quartz substrate and sample holder, and scanned using a Cu Kα radiation source. The diffraction angle (2θ) ranged from 5° to 60°, with a scanning rate of 10°/min.

2.3.4. Nuclear magnetic resonance

¹³C NMR spectra of the samples were acquired using an AVANCE III HD NMR spectrometer (Bruker, Germany) to investigate the solubility behavior and interaction mechanism of TSF and BSF in the IL. A total of 30 mg of EMIMAc (EAc) and EMIMAc/SF (EAc/TSF, EAc/BSF, EAc/TB0, EAc/TB450) was added to NMR tubes containing 0.6 mL of DMSO-d₆ and fully solubilized by moderate oscillation. The instrument operated at a frequency of 101 MHz, with DMSO-d₆ as an internal standard. The resulting NMR spectra were reported in ppm for chemical shift values.

2.3.5. Differential scanning calorimetry

The thermal properties of TSF/BSF composite samples prepared at various ultrasound power levels were analyzed using an STA 7300 synchronous thermal analyzer (TGA/DSC, Hitachi, Japan) in standard DSC mode. A fiber membrane sample (about 3 mg) was placed in a crucible, and then it was heated from the ambient temperature to 550 °C at 10 °C/min under nitrogen atmosphere for testing. The glass transition process was examined using a DSC7000X + RV differential scanning calorimeter (Hitachi, Japan) in temperature modulation mode (TM-DSC). In nitrogen atmosphere, the temperature of the sample was gradually increased to 250 °C at 5 °C/min. The modulation frequency and temperature amplitude are set to 0.02 Hz and 3 °C, respectively, so as to accurately control the temperature change. In order to ensure the accuracy of the measured data, we use aluminum and sapphire with known thermophysical parameters as standard samples for calibration.

2.3.6. Thermogravimetric analysis

The thermal stability of TSF/BSF composite samples was measured by TGA8000 thermogravimetric analyzer (PerkinElmer, USA). About 3 mg of the sample was heated from room temperature to 550 °C at a rate of 10C/min in a nitrogen atmosphere. The obtained thermogravimetric (TG) and first derivative thermogravimetric (DTG) curves provided detailed insights into mass loss and pyrolysis rate, critical parameters for evaluating thermal stability. The instrument was calibrated using standard reference materials such as aluminum and nickel, and each sample was tested in triplicate to ensure accuracy.

2.3.7. Nanoindentation

The AFM5000 II atomic force microscope workstation (Hitachi, Japan) was used to evaluate the mechanical properties of TSF/BSF composite samples. Before the test, the area function of probe tip is calibrated by using polycarbonate standard sample according to the standard method to reduce the influence of probe tip geometry on AFM test results. After the calibration, the indentation test of the composite membrane was carried out by using a spherical probe with a radius of 10 nm in tap mode. During testing, SF film samples were loaded and unloaded at a rate of 2000 nN/s, with a peak load range of 2800 nN to 5500 nN, maintaining the peak load for 0.5 s once reached. Repeat the test several times for each sample. The mechanical properties, including hardness (H), elastic modulus (E_s), and resilience (R), were calculated using formulas derived from our previous research [34].

2.3.8. Water contact angle

The DSA30S Water Contact Angle Meter (KRUSS, Germany) was used to determine the static water contact angle of TSF/BSF composite fiber membranes, assessing their surface hydrophilicity. A 2 μL drop of deionized water was placed on a uniformly thick and smooth sample membrane, and the contact angle between the droplet and the sample surface was recorded. To minimize errors, three measurements were taken at random locations on each sample, and the average value was calculated.

2.3.9. Cell viability

Mouse myoblasts (C2C12 cells) were used to evaluate the cell compatibility of TSF/BSF composite fiber membranes prepared by different power ultrasound. Place the sample with appropriate size at the bottom of 96-well plate, and use 75 % ethanol and ultraviolet radiation for 1 h to sterilize it in ultra-clean table. Subsequently, C2C12 cells were inoculated into well plates and cultured in a constant temperature incubator. After 6, 24 and 48 h, the culture medium in the well plate was removed, and 0.5 mg/mL thiazole blue (MTT) solution was added to continue the culture for 4 h. Then, the MTT solution was sucked away and dimethyl sulfoxide (DMSO) was added. Finally, EL-X800 microplate analyzer (BioTek, USA) was used to measure the absorbance to evaluate the cell activity. Each sample was tested with three parallel experiments and one control experiment. Cell viability was calculated using the following formula [35,36]:

$$\mathit{Viability}=\frac{D_s}{D_c}\times 100\%$$ (3)

where D_s is the absorbance of the sample group and D_c is the absorbance of the control group.

2.3.10. Biodegradability

The biodegradation experiment was conducted to evaluate the effect of varying ultrasound pretreatment on the degradation characteristics of TSF/BSF composite samples. Place the sample in a centrifuge tube, and add 5 mL deionized water, PBS buffer and 0.2 mg/mL protease solution. After that, it was placed in a constant temperature water bath at 37 °C to simulate the physiological environment to improve the activity of protease. After 1, 2, 3 and 5 days, the sample was washed with distilled water to remove residual protease and buffer. Then, SF film samples were dried in a vacuum dryer box until their weight was stable and there was no obvious change. The degradation of the SF fiber membrane was then calculated by comparing the weight before and after treatment. Additionally, the protease solution was replaced every 24 h to maintain catalytic activity and prevent experimental inconsistencies due to enzyme inactivation.

3. Results and discussion

3.1. Surface morphology

SEM, as an important tool for studying nanostructures, can reveal the microscopic morphology of material surfaces at the nanoscale, such as the surface roughness, cracks and other characteristics of nanostructures, which can directly impact their application performance [8,37]. As shown in Fig. 2a, the overall color of the fiber samples was brownish, with a visibly rough texture. This could be attributed to the irregular structural arrangement of the materials, resulting in the presence of nano-scale protrusions or depressions. Fig. 2b–f illustrate the SEM images of spray-spun SF fiber membranes treated with varying ultrasound power levels. These fibers exhibited a smoother surface, lacked significant defects or textures, and were arranged in an unordered manner, forming a complex network. This implied that ultrasound treatment enables the molecules to arrange in a more orderly manner, thereby reducing surface defects and resulting in a smoother surface. Moreover, due to differences in silkworm cocoon sources, silkworm species, and their rearing conditions, there was a significant variation in fiber thickness between Tussah silk fibers (TSF-0) and Bombyx mori silk fibers (BSF-0) [38]. As depicted in Fig. 2i and 2j, the fibers of Sample TSF-0 were relatively robust, with a diameter of 2.11 ± 0.24 μm, suggesting enhanced strength and toughness. In contrast, the fibers of Sample BSF-0 were finer, with a diameter of 0.84 ± 0.18 μm, imparting exceptional softness to the BSF.

Fig. 2.

 Physical photographs of SF composite fiber membranes (a); SEM images of fiber membranes prepared by ultrasonic spray spinning at different power levels: TSF-0 (b), BSF-0 (c), TB-0 (d), TB-150 (e), TB-300 (f), TB-450 (g), and TB-600 (h); as well as their average fiber diameter (i) and fiber surface area (j).

In the TB-0 sample, prepared via co-spray spinning of the two SFs, the fibers exhibited bent and twisted shapes with larger inter-fiber gaps. The fiber diameters also showed significant variability, primarily concentrated around 1.01 μm and 2.03 μm, corresponding to the diameters of individual BSF-0 and TSF-0 fibers, respectively (Fig. S1). These results indicate that the compatibility between the molecular chains of the two types of silk fibroin was limited during the co-spinning process, which hindered the orientation and crystallization of the silk fibroin molecules. However, the larger gaps between the fibers were beneficial for breathability and permeability. With increasing ultrasonic power, fiber size in the composite samples became more uniform. Notably, in Samples TB-450 and TB-600, compared to TB-150 and TB-300, the fiber count per unit area was significantly higher, with fibers more tightly interwoven and aligned, resulting in reduced inter-fiber gaps. The average fiber diameter decreased from 1.61 ± 0.65 μm (TB-0) to 0.96 ± 0.12 μm (TB-450), while the average fiber surface area increased from 237.39 ± 18.03 μm² (TB-0) to 310.85 ± 18.92 μm² (TB-450). The mechanical vibration and cavitation effects of ultrasound could enhance the mutual diffusion and mixing of silk molecular chains at the nanoscale, resulting in a more orderly arrangement of the molecular chains. This could lead to the formation of nanofibers with uniform dimensions. The increase in fiber quantity and their dense packing arrangement contribute to the construction of a more compact network structure at the nanoscale. Li et al. [39] confirmed that the ultrasonic cavitation could effective dispersion of agglomerates via micro-jet effects and modification of materials surface reaction kinetics through transient high-pressure microenvironments. Simultaneously, these structural modifications could significantly influence the mechanical properties, adsorption characteristics, and other attributes of the biomaterial. However, with further increases in ultrasonic power, fiber optimization began to diminish, leading to an increase in fiber diameter to 0.99 ± 0.14 μm in the TB-600 sample and a decline in fiber diameter distribution uniformity.

These results indicate that a moderate increase in ultrasonic power can reduce phase separation between the two proteins and promote uniform fiber thickness, optimizing surface characteristics and improving diameter distribution uniformity. This observation aligns with the findings of Jadaun et al. [40], who reported that high-power ultrasound treatment enhances fiber uniformity in dimension and crystallinity, increases the number of active groups on the fiber surface, and potentially improves surface hydrophilicity.

3.2. Structural properties

FTIR, XRD, and NMR were used to investigate the microstructure of materials [33,41]. As depicted in Fig. 3a, the characteristic absorption bands in the infrared spectrum of the SF composite membrane are primarily associated with its secondary structure, labeled as amide I (1710–1590 cm⁻¹), amide II (1590–1500 cm⁻¹), and amide III (1350–1200 cm⁻¹) [34,42]. When TSF was blended with BSF, the characteristic absorption peaks of sample TB-0 at 1630 cm⁻¹ and 3285 cm⁻¹ shifted slightly, suggesting a weak interaction between the two SFs. With increasing ultrasound power, the characteristic peaks of the composite samples underwent further changes. In the amide I band, the peak shifted from 1629 cm⁻¹ in TB-0 to 1621 cm⁻¹ in TB-450 and TB-600, while the peak in the amide II band shifted from 1508 cm⁻¹ to 1518 cm⁻¹. Moreover, increased ultrasound power sharpened the absorption peaks in the amide I and II bands, with the most pronounced effect observed at 450 W. These changes indicate that ultrasound treatment promoted intermolecular interactions within SF molecules, improving the stability and orderliness of their secondary structure [43,44].

Fig. 3.

Structural characterization of TSF/BSF composite fibrous membranes prepared by ultrasonic spray spinning at different power levels: FTIR spectra highlighting Amide I, Amide II, and Amide III as SF characteristic absorption bands (a); fitted plot of secondary structure content (b); XRD spectra (c); and ¹³C NMR spectra (d) of EMIMAc and EMIMAc/SF.

The amide I band is a key feature in protein infrared spectroscopy, primarily caused by C=O stretching vibrations in the peptide bond. It accurately reflects changes in the secondary structure of proteins [42,45,46]. Consequently, the amide I band of the TSF/BSF fiber membranes was fitted (Fig. S2) to assess the influence of ultrasound power on secondary structure. Fig. 3b and Table S1 show that in untreated samples, the β-sheet content of pure TSF (49.91 %) was slightly higher than that of BSF (48.55 %), a difference potentially linked to the adaptation of Tussah silkworms to their environment [47]. Sample TB-0 exhibited a β-sheet content of 50.76 %, suggesting that the TSF/BSF composite favors β-sheet formation. For samples treated with ultrasound, the β-sheet content followed the order TB-450 (56.49 %) > TB-600 (56.12 %) > TB-300 (55.01 %) > TB-150 (52.29 %), while the random coil content decreased from 23.02 % in TB-0 to 15.56 % in TB-450. These findings confirm that ultrasound treatment at optimal power (450 W) significantly enhances intermolecular hydrogen bonding and electrostatic interactions, improving structural order and promoting the transition from random coils to β-sheets.

XRD is a crucial analytical method for studying material crystal structures [48]. As shown in Fig. 3c, all samples exhibited similar diffraction patterns with weaker peak intensities and broad, disordered peaks. This pattern is attributable to the alternating distribution of crystalline and amorphous regions in silk and the absence of long-range ordered crystal structures [20,49,50]. Furthermore, in the XRD pattern of the pure SF sample, the diffraction peak at 24.83° was assigned to Silk I, while the peaks at 16.79° and 20.62° corresponded to Silk II (β-sheet) [51,52]. With increasing ultrasound power, the peak at 20.75° (TB-0) gradually shifted to 21.08° (TB-450), and the peak at 24.61° (TB-0) shifted to 23.94° (TB-450). Additionally, the peaks of TB-450 and TB-600 were very close to each other. These observations indicate that increasing ultrasound power facilitates the transition from Silk I to Silk II, enhancing the material’s crystallinity and stability.

The ¹³C NMR spectrum provides molecular structural information through chemical shift changes [41,53]. Interactions such as hydrogen bonding, ionic coordination, and van der Waals forces alter the chemical environment of atomic nuclei, affecting electron cloud density and reflecting these changes in the NMR chemical shifts [54,55]. Thus, analyzing carbon atom chemical shifts effectively assesses the interaction between TSF and BSF within the IL. The carbon atom numbering and chemical shifts of EMIMAc are shown in Fig. S3 and Fig. 3d. After dissolving 8.00 (w/w) SF, significant chemical shift changes were observed for specific carbon atoms (C2, C4, C7, C9, and C10) in EAc/SF (Fig. S4). The signals of C2 and C4 on the imidazole ring shifted to lower fields, likely due to their role as hydrogen bond acceptors forming hydrogen bonds with hydroxyl protons in SF, reducing the electron shielding effect and increasing the chemical shift [56]. The pronounced chemical shift changes of the carboxyl (C10) signal indicated strong hydrogen bond formation between Ac⁻ and hydroxyl groups in SF. This also influenced the electron distribution around C9, increasing electron cloud density and shifting the signal to higher fields [54,57]. Additionally, Table 1 shows that the absolute values of the chemical shifts (|△δ|) of the corresponding carbon atoms in EAc/TB0 were smaller than those in EAc/TSF. This may be due to the reduced proportion of polar amino acids (e.g., aspartic acid) in SF after BSF addition, leading to decreased interactions with EMIMAc and a lower electron cloud density. Ultrasonic treatment can induce conformational changes in SF, exposing more potential interaction sites [35,58]. Simultaneously, the cavitation and thermal effects of ultrasound strengthen hydrogen bonding and electrostatic interactions between SF polar residues and EMIMAc. Consequently, after 450 W ultrasound treatment, the |△δ| of the corresponding carbon atoms in EAc/TB450 was significantly larger than that in EAc/TSF.

Table 1.

Chemical shifts of specific carbon atoms in EMIMAc and EMIMAc/SF samples.^*

Sample	δ (ppm)
	C2	C4	C7	C9	C10
EAc	136.78	121.98	44.09	25.01	173.43
EAc/TSF	137.01	122.02	44.03	24.88	173.86
EAc/BSF	136.93	122.02	44.04	24.91	173.75
EAc/TB0	136.99	122.02	44.04	24.91	173.84
EAc/TB450	137.10	122.06	44.01	24.84	173.91

^*Specific C atoms are C2, C4, C7, C9 and C10 with significant chemical shift changes.

3.3. Thermal properties

Thermal analysis techniques provide a comprehensive understanding of the thermophysical properties of biomaterials, revealing the molecular dynamics and structural changes of polymers at different temperatures [59,60]. DSC and temperature-modulated DSC (TM-DSC) were employed to investigate the effect of ultrasound power on the thermal properties of TSF/BSF composite samples (Fig. 4a, 4b). As shown in Fig. 4a, the DSC curves of pure silk fibroin (SF) samples exhibited a single endothermic decomposition peak. TSF-0, with its compact molecular structure, displayed a higher peak temperature (T_p1 = 284.08 °C), whereas BSF-0 had a T_p2 of 249.74 °C (Table 2). Upon the addition of BSF to TSF, the composite sample TB-0 exhibited two endothermic peaks at 289.36 °C and 253.14 °C, corresponding to T_p1 and T_p2, respectively. This suggests that both SFs retain their chemical characteristics in the composite system, and the shift in peak temperatures indicates potential interactions, such as van der Waals forces and hydrogen bonding. With increasing ultrasound power, T_p1 and T_p2 shifted to higher temperatures, reaching their maximum at 450 W, confirming that increased ultrasound power enhances intermolecular interactions and raises the material's thermal decomposition temperature [29,44]. The TM-DSC curves (Fig. 4b) exhibited the glass transition temperature (T_g) and the heat capacity increment (ΔC_p) of the regenerated SF samples. Between 120 °C and 220 °C, composite sample TB-0 exhibited two glass transition temperatures (T_g^' and T_g), indicating phase separation between Tussah and Bombyx mori silk proteins within its structure. In contrast, the single T_g observed in the ultrasound-treated composite samples suggests that the components within the composite membrane were well-mixed, forming a homogeneous microstructure. [35,61]. This finding demonstrates that ultrasonic treatment effectively enhances the uniform mixing of SF, reducing phase separation in the material.

Fig. 4.

Thermal analysis curves of TSF/BSF composite samples prepared by ultrasonic spray spinning at different power levels: (a) DSC curve, where T_p refers to the peak temperature of thermal decomposition; (b) TM-DSC curve, where T_g is the glass transition temperature and ΔC_p is the heat capacity increment; (c) TGA curve, with inset (c') showing a magnified view; and (d) DTG curve, where T_pmax represents the temperature corresponding to the maximum mass loss rate.

Table 2.

Thermodynamic parameters of silk nanofiber membranes prepared by ultrasonic spray spinning at different power levels.^*

Sample	Tp1/°C	Tp2/°C	Tonset/°C	Tpmax/°C	ΔY/%	M550/%	Tg/°C	ΔCp/Jg−1°C−1
TSF-0	284.08		232.50	264.47	2.01	15.83	205.56	0.1085
BSF-0		249.74	225.86	260.78	2.62	14.06	198.31	0.1261
TB-0	289.36	253.14	233.41	265.93	1.58	17.25	130.27/206.58#	0.0964
TB-150	291.82	256.39	239.37	268.08	1.45	21.97	196.82	0.0810
TB-300	294.17	258.62	241.52	269.86	1.33	23.72	195.40	0.0539
TB-450	296.85	261.90	243.63	273.02	0.57	26.60	193.19	0.0372
TB-600	296.07	260.54	242.18	272.35	0.84	24.28	193.52	0.0413

^#The second T_g of TB-0.

^*Each sample was tested at least three times, with a numerical error within 2%.

Generally, a more ordered polymer structure corresponds to a higher T_g and a lower ΔC_p [62]. Accordingly, compared to BSF-0, samples TSF-0 and TB-0 exhibited higher T_g and smaller ΔC_p values due to their higher β-sheet crystal content (Fig. 3b). However, in ultrasound-treated samples, T_g decreased with increasing ultrasound power, from 196.82 °C in TB-150 to 193.19 °C in TB-450, with ΔC_p decreasing to 0.0372 Jg⁻¹°C⁻¹ (Table 2). This demonstrates that the influence of ultrasound treatment is intricate and multifaceted: it promotes uniformity in the local structure of the composite, reducing heterogeneity caused by phase separation. Simultaneously, high-power ultrasound may induce structural flexibility in the non-crystalline regions of SF and increase the exposure of polar groups [63,64]. This synergistic effect collectively contributes to the observed decrease in T_g at the macroscopic level.

TGA further elucidated the impact of ultrasound power on the thermal stability and weight loss behavior of the SF composite samples. As depicted in Fig. 4c and 4d, all samples exhibited minor weight loss between 30 °C and 150 °C due to the evaporation of free and bound water in SF. The pure Bombyx mori silk fibroin sample (BSF-0) had the highest water content at 2.62 %, which decreased to 1.58 % upon blending with TSF. Between 200 °C and 350 °C, increasing temperature weakened intermolecular forces and caused peptide bond cleavage, leading to rapid mass loss [8,65]. Above 400 °C, decomposition continued at a slower rate. The thermal decomposition profiles were consistent with the DSC analysis results. The pure TSF and BSF samples began to decompose at 232.50 °C and 225.86 °C, respectively, while the composite sample TB-0 had a slightly higher onset temperature (T_onset) of 233.41 °C. With increasing ultrasound power to 450 W, TB-450 exhibited the highest T_onset at 243.63 °C. Additionally, the temperature at which the maximum rate of weight loss (T_pmax) occurred increased from 265.93 °C in TB-0 to 273.02 °C in TB-450 (Table 2). These findings suggest that interactions such as hydrogen bonding and van der Waals forces between TSF and BSF contribute to enhanced thermal stability [62,66]. Furthermore, ultrasound treatment at an optimal power level facilitates the transition of SF to a more stable β-sheet structure, thereby improving the sample's thermal resistance [67,68].

In summary, the results from DSC and TGA indicate an intermolecular interaction between two components in the composite fibers. Ultrasound enhances this interaction, promotes uniformity of fibers, reduces phase separation in the composites, and facilitates a more stable β-sheet structure, which improves the thermal stability of the material.

3.4. Mechanical properties

The mechanical properties of biomaterials, such as hardness, elastic modulus, and resilience, are crucial for evaluating their potential in tissue engineering and can also reflect the compactness and order of molecular structures [34,69]. Fig. 5a presents the force-distance curves for selected composite fiber membranes, with sample TB-450 exhibiting the highest peak load (5470 nN), indicating a high load-bearing capacity. Fig. 5b and 5c show the hardness (H) and elastic modulus (E_s) of the TSF/BSF composites. Compared to the untreated pure silk fibroin films TSF-0 and BSF-0, the composite TB-0 exhibited higher H (10.39 GPa) and E_s (0.533 GPa). After ultrasound treatment, particularly at 450 W, both H and E_s reached their maximum values, increasing by 1.68 and 1.63 times compared to TB-0, respectively. However, when ultrasound power was increased to 600 W, a decline in H and E_s was observed.

Fig. 5.

Force-distance curves (a) for composite fiber membranes TB-0 and TB-450, and hardness (b), elastic modulus (c), and resilience (d) of TSF/BSF composite samples prepared by ultrasonic spray spinning at different power levels, *p < 0.05, **p < 0.01.

The resilience of the SF composite samples is shown in Fig. 5d. The addition of BSF to TSF caused a slight decrease in resilience to 69.2 % in TB-0. However, with increasing ultrasound power, the resilience of the TSF/BSF composites gradually improved, reaching a maximum of approximately 93.0 % at 450 W and 600 W. These results indicate that although incorporating BSF into TSF slightly reduced resilience, it enhanced the material's ability to resist deformation under load. This is likely due to interactions between the two SFs, which facilitate effective cross-linking of molecular chains and the formation of β-sheet structures. Furthermore, ultrasonic treatment at an appropriate power (450 W) significantly enhances the hardness, elastic modulus, and resilience of the TSF/BSF composite film. This improvement is attributed to the increased uniformity of the TSF and BSF mixture induced by ultrasound, which strengthens intermolecular interactions, improving the composite's structure and ultimately leading to enhanced stability and mechanical strength [55,70].

3.5. Biological properties

In regenerative medicine, a material's hydrophilicity, cytocompatibility, and biodegradability are crucial for its viability in biological applications [36,71,72]. Hydrophilicity not only regulates interactions with the biological environment but also influences cell adhesion and proliferation. Fig. 6a illustrates the change in water contact angle for TSF/BSF composite fiber membranes. The water contact angle increased slightly from 72.5° (TSF-0) to 73.8° (TB-0) upon the addition of BSF, which contains fewer hydrophilic amino acids (e.g., aspartic acid) than TSF. With increasing ultrasound power, the water contact angle decreased to 49.7° (TB-600), significantly enhancing surface wetting. This improvement in hydrophilicity may be attributed to ultrasound-induced uniform mixing of SF within the composite and the rearrangement of molecular chains, which expose more polar groups (e.g., hydroxyl and carboxyl) on the surface, thereby increasing interactions with water molecules [64,73]. The morphology of SF composite fiber membranes also influences their hydrophilicity. A compact fiber arrangement increases the contact area, allowing more water molecules to interact with the fiber surface. Additionally, the presence of micropores or capillaries within the membrane facilitates water adsorption and penetration through these channels [74,75], further enhancing the overall hydrophilic character of the membrane.

Fig. 6.

(a) Water contact angle of TSF/BSF composite fibrous membranes prepared by ultrasonic spray spinning at different power levels; (b) viability of C2C12 cells after 6, 24, and 48 h of incubation on TSF/BSF fibrous membranes, detected by the MTT assay (*p < 0.05, **p < 0.01); and degradation profiles of composite membrane samples in (c) PBS buffer and (d) protease solution.

Fig. 6b demonstrates the proliferation of C2C12 cells on SF composite materials. Across 6, 24, and 48-hour culture periods, cell viability in all experimental groups was higher than that of the control, indicating a promotional effect of the SF samples on cell proliferation. Notably, the cell viability of the unsonicated composite sample TB-0 was lower than that of TSF-0 at the same culture time, possibly due to the decreased hydrophilicity after composite formation, which affects cell growth and proliferation. In contrast, ultrasound-treated SF composite samples exhibited higher cell viability, which correlated positively with ultrasound power. At 6 h, the cell viability of TB-150 was 111.29 %, a 3.48 % increase compared to TSF-0. With increasing ultrasound power, the cell viability of TB-450 and TB-600 further increased to 121.26 % (6 h) and 122.47 % (6 h), respectively, indicating a significant enhancement. After 48 h of cultivation, cell viability increased even more, reaching 159.33 % (TB-450) and 163.72 % (TB-600). These findings suggest that SF materials possess excellent biocompatibility. Ultrasonic treatment can modify the structure and properties of silk fiber composites by promoting the formation of micro- and nanoscale structures, increasing β-sheet content and surface-active groups, and enhancing both hydrophilicity and mechanical properties [43,76,77]. Consequently, composite samples treated with high-power ultrasound are more effective in promoting cell adhesion and proliferation.

Degradation studies are critical for assessing and optimizing the biodegradability of biomaterials, ensuring their safety and functionality in biomedical applications [71]. To investigate the effect of ultrasound power on the degradation characteristics of SF composites, samples prepared under varying ultrasound powers were immersed in deionized water, PBS buffer, and protease solutions, with the corresponding degradation curves shown in Fig. S5 and Fig. 6c, 6d. Under PBS immersion for the same duration (Fig. 6c), the degradation rates of the samples followed the order: BSF-0 < TB-0 < TSF-0 < TB-150 < TB-300 < TB-450 < TB-600. It was observed that the degradation rate of composite sample TB-0 was lower than that of the unsonicated pure Bombyx mori silk sample TSF-0. However, as ultrasound power increased, the degradation rate of the composite samples also increased. Additionally, the degradation behavior in deionized water followed a similar trend to that in PBS (Fig. S5) but with lower overall degradation rates. In contrast, in protease solution, degradation was significantly accelerated due to the enzymatic hydrolysis of SF. Specifically, in the initial degradation phase, the degradation rate of TB-0 was 12.76 %, which increased to 23.09 % for TB-600. After 5 days of continuous degradation, the degradation rate of TB-600 reached a maximum of 60.32 % (Fig. 6d). As reported in the literature [5,13,78], SF degradation performance is influenced by factors such as molecular structure, surface morphology, and hydrophilicity. Li et al. [13] found that the degradation rate of non-domesticated silk fibroin (ASF) was faster than that of Bombyx mori silk (BSF), which was attributed to ASF’s lower molecular weight and the formation of more chemical crosslinks with higher active groups. Lu et al. [78] discovered that SF degradation is not only related to crystal content but also interacts with hydrophilicity. They found that SF films with the highest β-sheet content exhibited the highest degradation rate, and SF samples with stronger hydrophilic interactions degraded faster in enzymatic solutions.

3.6. Mechanism of composite fiber formation

Blending is an effective method for fabricating protein composites with tunable mechanical properties, degradability, biocompatibility, and functionality by precisely controlling temperature, protein concentration, and ultrasonic treatment [8,13,35]. Ionic liquids (ILs), valued for their environmental sustainability and versatile physicochemical properties, are widely used in biomass processing, facilitating efficient SF dissolution and regeneration [15,18,19]. In silk fiber processing, ILs streamline SF regeneration and material formation by minimizing processing steps. Fig. 7 illustrates the interaction of imidazolium cations (EMIM⁺) and acetate anions (Ac⁻) in EMIMAc with SF functional groups [79,80]. EMIM⁺ forms hydrogen bonds with SF hydroxyl (–OH), amino (–NH₂), and amide (–CONH–) groups, while its alkyl chain interacts hydrophobically with nonpolar amino acid residues. Conversely, Ac⁻ forms electrostatic interactions with SF’s carboxyl (–COOH) and cationic amino (–NH₃⁺) groups, disrupting intramolecular hydrogen bonds and enhancing TSF/BSF dissolution in EMIMAc [34,81]. Due to differences in amino acid composition and molecular structure, TSF and BSF exhibit poor compatibility, leading to phase separation at the microscale [82,83]. Hydrogen bonding, hydrophobic interactions, and van der Waals forces within TSF and BSF chains differ from those between heterogeneous molecules, forming distinct microphase structures. Variations in solubility and crystallization properties further contribute to phase separation, as thermodynamic incompatibility drives the system to lower its free energy through phase separation mechanisms. The presence of two glass transition temperatures in sample TB-0 confirms this phenomenon.

Fig. 7.

Mechanism of ultrasonic regulation of TSF/BSF composites in EMIMAc.

Ultrasound treatment plays a pivotal role in enhancing silk protein mixing and reducing phase separation in the EMIMAc system. The cavitation effect generates localized high temperature and pressure, disrupting pre-existing hydrogen bonds in TSF and BSF, thereby improving dispersibility in ILs [39,84,85]. Additionally, mechanical shearing forces further break and reorganize silk protein chains, altering their conformation, increasing β-sheet content, and exposing surface-active groups [58,86]. These modifications strengthen intermolecular interactions, reduce aggregation, and lead to a more uniform microstructure, effectively minimizing phase separation. Ultrasound also enhances silk protein solubility, facilitating homogeneous dispersion and reducing phase separation due to solubility differences [34,87]. Moderate-power ultrasound promotes the formation of uniform fiber sizes and interwoven Ran structures, strengthening intermolecular hydrogen bonding and electrostatic interactions, while improving protein crystallinity, orderliness, and overall composite properties. Specifically, treatment with 450 W ultrasound resulted in the highest β-sheet content (56.49 %), an increase in T_onset by 10.22 °C, and E_s and cell viability enhancements of 1.63-fold and 1.19-fold, respectively. Overall, appropriate ultrasound treatment significantly enhances the structural uniformity, thermal stability, and mechanical properties of TSF/BSF composites, reducing phase separation of different silk proteins and yielding highly stable silk composite fiber membranes.

4. Conclusion

In this work, we achieved TSF and BSF miscibility via ultrasonic-assisted spinning, using ionic liquids (ILs) as a medium, and successfully fabricated TSF/BSF composite fiber membranes with enhanced morphology and structural properties. SEM analysis showed that untreated samples had uneven fiber thickness and large gaps, which improved with increasing ultrasound power, leading to more uniform fiber diameters and a tighter weave. Microstructural studies (FTIR, XRD, and ¹³C NMR) revealed weak interactions between TSF and BSF, while moderate ultrasound treatment (450 W) enhanced inter-protein interactions, reduced aggregation, and minimized microphase separation. Ultrasound also improved crystallinity and molecular order, significantly influencing structural stability and material properties. Thermal analysis and nanoindentation results indicated that ultrasound-treated SF composites exhibited enhanced thermal and mechanical stability, with 450 W and 600 W yielding optimal biological performance. Water contact angle, cell viability, and degradation experiments showed that ultrasound reversed the hydrophilicity and biocompatibility decline post-composite formation. Additionally, ultrasound improved silk fibroin solubility and dispersion in EMIMAc, ensuring uniform blending and preventing microphase separation due to solubility differences. These findings highlight the role of ultrasound in tuning SF composite properties, offering insights for developing high-performance biomaterials with applications in biomedicine and material science. Such applications include tissue scaffolds, drug carriers, targeted drug delivery systems, wound dressings, and biological sensors. Nevertheless, current research presents certain limitations. Notably, there is a deficiency in long-term studies on the effects of nanofibers on cell viability, which could offer more profound insights into their applications. Future research will address this gap by supplementing cell cycle assessment experiments and establishing biological models. Additionally, we will explore specific practical application scenarios in greater detail to provide comprehensive theoretical support for the fabrication of silk-based and other natural biomaterials [88,89,90,91], thereby enhancing their potential clinical applications.

CRediT authorship contribution statement

Xincheng Zhuang: Writing – original draft, Investigation, Formal analysis, Data curation. Weiting Gong: Data curation. Fang Wang: Writing – review & editing, Writing – original draft, Supervision, Funding acquisition, Conceptualization. Xiao Hu: Writing – review & editing, Funding acquisition, Conceptualization.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary data

The following are the Supplementary data to this article: