Triaxial electrospinning process
The clear preparation process of the triaxial electrospun fibre was based on the dyeing of the working fluids. The inner working fluid was dyed dark blue by methylene blue, whereas the middle working fluid was dyed red by magenta. In this work, inner and outer working fluids were driven by the same lateral driving force, and the middle working fluid had a vertical driving force due to the internal structure of the homemade spinneret (Fig. 2a–d). The vertical collection method was designed to facilitate the observation and preparation of fibres. Similarly, the F1 fibres prepared by modified coaxial electrospinning also used this method (Additional file 1: Fig. S1). The inner working fluid was dyed dark red by magenta, whereas the outer working fluid was transparent. Although a small amount of dye was diffused into the transparent sheath, the core–shell structure could still be distinguished by the shade of colour. For the modified triaxial electrospinning process, Fig. 2e shows the compound droplets with different colours around the homemade spinneret caused by different driving forces. At the beginning stage, the blue liquid was surrounded by a transparent liquid in Fig. 2e (I) under the driven of inner and outer working fluids. The red dyeing fluid gradually appeared in the middle of the droplet driven by middle working fluid. Increased driven speed of the middle working fluid resulted in red area in the droplet. In Fig. 2e(II–VI), the compound droplet exhibited a transparent–red–blue structure from outside to inside, including the dripping of the droplet.
Figure 2f showed a clear fabrication process of the electrospun triaxial fibres, where the voltage was added after the three colors appeared in the droplet. The three main stages in the fabrication process of fibres included compound Taylor cone, straight jet area, and a complex circle whipping area. The whole process exhibited a magic phenomenon of the preparation of multi-layers fibres in a short time. The partial zoom of the compound Taylor cone in the right corner of Fig. 2f, including a blue inner layer, a red middle layer, and a transparent outer layer.
Surface morphology and the internal structure of fabricated fibres
The microscopic surface morphology of fabricated fibres is shown in Fig. 3. F1 fibres exhibited a uniform and smooth cylindrical surface with an average diameter of 500 ± 130 nm. By contrast, F2–F4 fibres prepared by modified triaxial electrospinning did not have perfect morphology and showed a cylindrical shape with a certain number of bead-on-a-string fibres. The average diameter of F2–F4 fibres was slightly increased compared with that of F1 fibres, which implied the formation of a double–layer structure. Few beads-on-a-string morphologies in F2–F4 fibres could contribute to the PEG component, which could not be optimized by using the modified triaxial electrospinning. Meanwhile, the average diameter of F2–F4 had no significant connection with the PEG content from SEM images.
As shown in Fig. 4, the internal structure of the fabricated fibres was observed clearly by TEM. F1 fibres exhibited a straight and single-layer structure, while the core–shell structure of F2–F4 fibres was apparent. However, the core layer was not exactly in the centre of the fibre in Fig. 4b–d. This result was probably attributed to the randomness of the preparation of the fibres in the electrospinning process. As shown in Fig. 4c, d, the fibre with beads-on-a-string morphology exhibited a core–shell ‘bubble’ structure, which was closely related to the electrospinning process. Linear, spindles/beads-on-a-string, and particle morphology produced in the electrospinning process were normal, which were inseparable from the elasticity and viscosity of the working fluid. In this work, the incorporation of the PEG component may decrease the viscosity of F2–F4 working fluids. The rapid stretching effect of electrostatic action on F2-F4 working fluid is reduced during the bending and whipping area in Fig. 2f. Thus, the incomplete stretching of the working fluid under the electric field leads to linear with beads-on-a-string morphology (‘bubble’ structure).
The different greyscales in black and white TEM images suggested the structural changes, and the grey level could be influenced by thickness, density, and elements . In this work, CA was the principal component in the two prepared fibres, and the PEG component in F2–F4 could be ignored due to the few contents. Thus, density and thickness were the key factors that led to the grey level. In Fig. 4b–d, 2% drugs were stored in the core of F2–F4 fibres, which increased the molecular entanglement of the CA polymeric matrix to a certain degree, thus increasing the core density of F2–F4 fibres. In addition, based on the top–down observation method in the TEM, the grey level of the core layer was composed of the original core layer and the part of the sheath layer, which led to a ‘thick’ core layer. Therefore, the synergy of density and thickness led to a high grey level of core in Fig. 4b–d.
XRD and FTIR
In Fig. 5a, a small number of short peaks appeared in the X-ray diffraction (XRD) pattern of CA powder (the main component in fibres), which suggested that CA powder had a certain crystallinity. However, these early short peaks did not appear in the XRD pattern of the four fabricated fibres in Fig. 5a. The CA component was evenly distributed after the process of electrospinning, and the substance with a certain degree of crystallinity no longer existed. In the XRD pattern of PEG, the highest peak shown at 23.7° represented the (112) plane, and peaks around 19.1° and 27.3° were assigned to the (120) and (222) planes, respectively. These sharp and other low intense peaks indicated that PEG was a typical crystal material [43,44,45]. For F1–F3 fibres, only one hump appeared in the corresponding XRD patterns, indicating amorphous forms. The low PEG component had no negative effects on the physical form of F2 and F3 fibres, but the XRD pattern of F4 fibre containing 6% (w/v) PEG showed a faint peak near 13–25°. For Cur, a series of sharp peaks also responded to typical crystal materials. By contrast, F1–F4 fibres had no characteristic peaks in their XRD patterns, which suggested that the physical form of drug crystal changed after the electrospinning process (from a crystalline state to an amorphous state). Electrospinning can be regarded as a physical drying process, where working fluids are converted into solid fibres during solvent evaporation. Prior to electrospinning, all working fluids were homogeneous transparent solutions. Drug molecules were highly dispersive to the surface of the solid micro/nanofibre in a short time during the electrostatic drying process and the crystal structure of the drug transferred to an amorphous state. Although the drug concentration in the fibre increased, no sharp peaks appeared in the XRD pattern of the drug (Additional file 1: Fig. S2). The electrospinning process is a reliable process to transform the drug from a crystalline to an amorphous state.
The FTIR spectra of raw materials (i.e., PEG, CA, and Cur) and fibres are shown in Fig. 5b. For PEG, the characteristic peaks at 2884 cm–1 indicated the asymmetric bending vibration of methylene groups (-CH2) . In addition, the characteristic absorption bands at 946 and 841 cm–1 were related to the asymmetric stretching vibration of ether groups (-C-O-) [44, 47]. Although characteristic peaks of PEG did not appear in the spectrum of F2 and F3 fibres, a higher PEG content in F4 fibres showed clear characteristic peaks at 2876 cm–1 in the FTIR spectrum. Cur exhibited a carbonyl (C = O) at 1626 and 1602 cm–1 and olefins (-C = C-) at 1506 cm–1 [48,49,50]. F1–F4 fibres had the characteristic peaks of carbonyl and olefins of Cur, which proved that Cur had been successfully encapsulated. In addition, hydroxy groups (PEG) and free carbonyl groups (Cur and CA) would form the hydrogen bond to increase the compatibility in fibres (Fig. 5c), such as providing a stable exist for drug molecules in fibres.
Water contact angle and swelling performance
The hydrophilicity of the drug-loaded fibres conducted by the water contact angle test is a key point in a rapid drug delivery to the desired site of the patient. The first research between water contact angle and interfacial tension of the drops on a plane was proposed by Thomas Young in 1805 . Prior to preparing the functional hydrophilic materials, three hydrophilic polymers (i.e., PVP K13–18, PVP K30, and PEG) were added to the CA working solution. The addition of PVP hardly affected the hydrophilic properties of fibres, but the PEG component almost instantaneously improved the hydrophilicity of CA (Additional file 1: Fig. S3). The surface morphology of the fibres after adding hydrophilic ingredients hardly changed before and after water intrusion (Additional file 1: Fig. S4). This result indicated that the added agents PEG and CA polymeric matrix formed uniform fibres after the electrospinning process. In other words, the PEG component was evenly distributed in the fibres and did not form the pore structure after the water invade. The hydrophilicity of CA drug carriers could be adjusted by controlling the content of PEG. When the PEG content reached 3%, the water molecule was ‘absorbed’ by fibres in a short time. Thus, fibres showed a highly hydrophilic behavior. However, when the PEG content was reduced to 1%, fibres exhibited a hydrophobic state in the first 10 s (water contact angle > 90°), and fibres instantly changed from hydrophobic to highly hydrophilic after 10 s (Additional file 1: Fig. S5).
Triaxial electrospinning was used to fabricate functional F2–F4 fibres on the basis of the above acknowledge, where the CA polymeric containing PEG was designed as a middle layer to improve the hydrophilicity. In Figs. 6a, the water contact angle of F1 fibres was kept around 120° in 140 s, as a stable hydrophobic state. As time went by, the hydrophilic property of F1 fibres was enhanced, where the water droplet was gradually ‘absorbed’ by F1 (Additional file 1: Fig. S6). By contrast, F2–F4 fibres showed a unique dynamic water contact angle process in Fig. 6a. With the increase of content of PEG in core–shell fibres, the short–term hydrophilicity of the fibres was greatly improved. For F2 fibres, the water droplet was smaller at 120 s than that of the original state (from 122.47° ± 1.75° to 85.28° ± 3.56°). In the following 20 s, F2 fibres gradually exhibited a highly hydrophilic state. For F3 and F4 fibres, the time needed for this process was less than 5 s, and F4 with higher PEG content exhibited a water contact angle approaching 0° (5.37° ± 2.15°) than 20.12° ± 6.03° of F3 fibres. Compared with those of F1 fibres, the dynamic water contact angle images of F2–F4 fibres showed a transition from hydrophobic to hydrophilic in a short time, and the time could be manipulated by amount of PEG.
In Fig. 6b, a small amount of hydrophilic PEG in the fibres ‘absorbed’ the water molecules for the first time when the water came into contact with the surface of fibres. The molecular micropores left by dissolved PEG component formed a ‘passage’ in a short time to guide the water flow inward, which accelerated the process of invading the fibre matrix. Thus, water droplets decreased in size over time. The good hydrophilicity resulting from the combination of materials and structure will affect its other functionality, such as swelling properties and drug release properties of F2–F4 fibres.
Figure 7 showed a dynamic water uptake efficiency of the prepared four fibres. F2–F4 fibres, prepared by modified triaxial electrospinning, had better water storage capacity than F1 fibres in the first 20 min. Specially, F3 and F4 fibres only needed 5 min to reach saturation, while F2 fibres needed 0.5 h. The increased PEG content had improved the water uptake efficiency of fibres at the early stage, which was similar to water contact angle results. By contrast, F1 fibres could absorb water at a considerable rate until 90 min, then stabilised at 120 min (near 7555%). The lower water uptake efficiency of F2–F4 than that of F1 suggested that the ‘passage’ left by the dissolved PEG component caused the water loss to a certain extent. Gap left by the dissolution of PEG reduced the water–holding capacity of functional F2–F4 fibres. The macrophotographs before and after the experiment (Additional file 1: Fig. S7) were collected to analyse the distinct swelling phenomenon of F1 and F2 fibres. F1 fibres were ‘fluffier’ and filled with many thin fibres. F2 fibres were relatively ‘dense’, in other words, their density was relatively high, which was similar to TEM results. In addition, F2–F4 left less fibrous residue on the aluminum foil than F1 (Additional file 1: Fig. S8). This result contributed to the core–shell structure produced using triaxial electrospinning, which led an increase in the bonding force between fibres. Thus, F2–F4 fibres could be easily removed from the aluminium foil. ‘Fluffy’ F1 fibres provided a strong ability to store enough water, whereas ‘dense’ (not enough space) F2–F4 fibres led strong shrinking phenomenon.
Moreover, the fibres’ swelling properties between recent articles and this work were compared (Table 2). Functional F2–F4 fibres had similar or even better swelling efficiency compared with other fibres as we know. In this work, although the swelling efficiency of F2–F4 fibres were weaker than that of F1 after 2 h, the initial ability of water uptake had remarkable potential in special situations, such as producing novel drug delivery carriers.
The electrospun fibre was the high-efficiency drug-loaded carrier [57, 58]. In this work, the calculated drug loading efficiencies of F1–F4 fibres were 100.79% ± 0.55%, 96.81% ± 4.45%, 93.16% ± 5.10%, and 92.09% ± 3.04, respectively. The high loading efficiency was due to the electrospinning nanotechnology and CA polymer. CA was proven to be an excellent biomaterial in drug delivery systems in recent decades [59,60,61]. The good hydrophilic and undissolved properties of CA brought a slightly longer drug dissolution behaviour. In Fig. 8a, monolithic F1 fibres had a 48 h drug dissolution profile, whereas functional F2 fibres had a more than 96 h profile. The middle blank layer in the triaxial electrospinning process provided a ‘barrier’ to prevent the process of drug release from inside out and prolonged the drug release time. However, this blank layer lost its ‘barrier’ effect in F3 and F4 fibres due to the high PEG content (over 3%), drugs were completely released in 12 h. Tailing–off effect is inevitable for drug carriers at the late drug release stage and is useless for patients. Fortunately, F3 and F4 fibres had improved drug release profiles, which eliminated the tailing–off effect and made fully use of drugs. The dissolution profile of Cur powders was irregular, and only 40.92% ± 2.03% of drugs were dissolved after the long term shaking time. The total drug release from the four fibres were 93.57% ± 2.74%, 87.54% ± 2.98%, 84.21% ± 3.10%, and 80.76% ± 6.24%, respectively. High drug dissolution suggested that electrospun fibres could improve and prolong these kinds of water-insoluble drugs. The low total drug dissolution of F2–F4 fibres could be attributed to the ‘barrier’ effect of the core–shell structure, which prevented the diffusion of the last few amounts of drugs.
Figure 8b shows the relationship between time and relative drug release. F1, F3, and F4 fibres exhibited a similar time required to reach each release stage, while F2 needed a long time period. With increased drugs released, the gap between the time required for F2 and other fibres was widened. In other words, the long time needed to reach the set release amount meant a slow drug release. These statistics provided direct evidence to demonstrate the improved sustained–release properties of functional F2 fibres with 1% PEG.
First-order and Peppas models were established to evaluate the drug release mechanisms in Fig. 8c and 8d [62,63,64]. All fibres were perfected following the first-order model due to good regression results (R12 = 0.9981, R22 = 0.9923, R32 = 0.9768, and R42 = 0.9777). The value of n in the Peppas model is the key parameter for evaluating release mechanisms. n values in formulas were less than 0.5, which indicated that drug release was dominated by the Fickian diffusion. Thus, all fibres had similar diffusion mechanisms. In addition, the R22 = 0.9535 of F2 was higher than that of other fibres (R12 = 0.8858, R22 = 0.8905, and R32 = 0.8721), which suggested that the regression result of F2 was better than that of other fibres. The above results reflected that a suitable combination of structure and materials had a positive effect on the drug release mechanisms, such as the suitable additive of hydrophilic PEG and the core–shell structure in this work.
Drug release mechanism
The drug mechanisms of fibres were predominantly affected by polymeric properties (such as hydrophilic property) and structure. CA, as a natural swelling material after ‘absorbing’ water, has a high water-uptake performance that is beneficial for drug release. For monolithic F1 fibres in Fig. 9, drug molecules were diffused by water molecules into the release medium when they invaded the polymeric matrix, and the drug concentration difference between inside and outside the fibres was directly proportional to the drug release rate. The molecular gap left by drug molecular increased with the swelling of CA, and the high remaining water efficiency of drug carriers F1 provided favorable conditions for drug diffusion into the outside release medium. By contrast, the drug release mechanism of F2–F4 was complicated. Generally, the wettability of drug carriers had a remarkable influence on drug release, and excellent wettability always leads to a rapid drug release process. The incorporation of PEG had improved the hydrophilicity of functional F2–F4 fibres and the hydrophilicity improvement effect was significant when the content reached 3%. Thus, F3 and F4 had a quicker drug release that F2. Compared with the relatively quick release of F1, the continuous sustained-release process of F2 in the later stage could be attributed to the structural design and PEG component of fibres. Although the PEG component was dissolved as the water entered, the water-insoluble CA matrix extended the distance of water molecules to the core layer, as shown in Fig. 9. The blank sheath of F2 was designed to delay the water molecules arriving at the core drug storage. Furthermore, the swelling performance of F2 could also be positively affected by the structural design, the outer layer swelled first and delayed the water invasion. Meantime, molecular gaps, left by drug molecules and dissolved PEG components, would be increased during the swelling process. For drug release, the increased molecular gaps after water invade swelling fibres weakened the protective structure of the drug stored in core layer, which accelerated the released rate. In the later stage II, drug molecules passed through the swelling blank sheath to the external release medium under the diffusion effects. However, more molecular gaps (after PEG dissolved) in F3 and F4 fibres could ‘absorb’ enough water molecules to invade fibres, which greatly weakened the role of the core–sheath structure in preventing water intrusion. This design of core–sheath fibres with a high PEG component accelerate the drug release to deal with the tailing–off effect.
The improvement of materials and structure brought the expected drug-release performance. The corporation of hydrophilic PEG improved the hydrophilicity of fibres and accelerated drugs release. Core–sheath structure provided a ‘barrier’ layer for drug reservoir to prolong the drug release. In this work, the ingenious combination of the two designs could not only improve the hydrophilic properties of fibres, but also greatly improved the sustained release performance. Simple preparation of functional materials by using the ‘one–pot’ modified triaxial electrospinning nanotechnology had a remarkable potential in pharmaceutics, drug delivery systems, and environment area. Based on the combination model of the structure and materials in this work, a considerable number of new strategies could be exploited, and applications were wide.