Low molecular weight chitosan (LMWC), (15 kDa, 100%DDA) was prepared in-house using the acid hydrolysis method by the use of 2M Hydrochloric acid. The produced chitosan was characterized for its degree of deacetylations by the 1^st derivative UV method, the viscosity average molecular weight of chitosan was determined using Marck-Houwink Equation. Poly (lactic acid) (PLA) (3052D), poly (caprolactone) (PCL) (23 kDa), and poly (vinyl) alcohol (PVA) (30kDa) which were used as a copolymer for the nanofiber preparation were purchased from Unic Technology Ltd, Taiwan. On the other hand, dichloromethane (99.8%) was purchased from Merck, Germany.

About 9 g of PVA was dissolved in water to prepare a solution of 9 wt% concentration. Then, specific amounts of PCL, and PLA were dissolved in dichloromethane or in the (30:70) mixture of (acetic acid: formic acid) to prepare solutions of 3 wt% PCL, 10% PCL, and 6 wt% PLA. The 15 kDa chitosan was blended with the previously prepared PVA, PCL, and PLA solutions by adding the required amounts of chitosan and kept them under continuous stirring for 24 h. The composition of the different grades of the prepared spinning blends are shown in Table 1.

The electrospinning of chitosan blended nanofibers was conducted at room temperature using Nanospider Electrospinning Machine (Elmarco) (which is shown in Fig. 1) by applying a voltage of 45 kV at a carriage speed of 200 mm/sec, during which the relative humidity in the spinning chamber was kept in the range of 60–69%. The electrospun fibers were collected using a take-up cylinder, on a conveyer which is placed between the two wire electrodes, and at a height of 210 mm from the lower wire.

Figure 1. 

Schematic representation of the wire electrospinning technique.

Morphology of the produced fibers was observed through SEM (Hitachi Tabletop TM 3030, Japan). The samples were placed on aluminum stubs after sputtering with a layer of gold, and the diameters of the electrospun fibers were calculated using ImageJ 1.52a software (National Institutes of Health, USA). The average fiber diameters, and diameter distributions was obtained by taking measurement of 100 fibers selected from 3 different samples.

The FT-IR analysis is based on the identification of absorption bands associated with the vibrations of functional groups within macromolecules. Herein, dried samples were analysed by taking FTIR spectra in the range of 400 cm^-1 to 4000 cm^-1 using a spectrometer (Thermo Scientific Nicolet IS50).

The DSC analysis is commonly used for the measurement of heat effects on polymers. Generally, this method is used for detecting phase changes such as crystallization, melting, glass transition, and decomposition. In this study, a Perkin Elmer, DSC8000 instrument was used for the DSC analysis. About 2 mg of each powdered sample was placed in closed platinum pans and continuously heated from 30 °C to 300 °C at 1 °C/min.

Chitosan is well known to be among the most naturally abundant biopolymers. However, the strong inter- and intramolecular interactions between the amine groups of chitosan often hinder sufficient chain entanglement. This is a major challenge to prepare chitosan nanofibers with smooth morphology. Previously, the spinnability of chitosan has been reportedly improved by incorporating certain co-spinning agents, including different types of synthetic or natural polymers (Sangsanoh et al. 2006; Mendes et al. 2016). Therefore, different trials have been carried out to prepare chitosan nanofibers.

Herein, chitosan nanofibers were prepared through free surface wire electrospinning process, which is considered a new technique with fewer studies reported. Hence, during the trials, this study relied on previous studies reported about chitosan nanofibers using needle electrospinning process (Haider et al. 2009; Kriegel et al. 2009; Ziani et al. 2011; Koizumi et al. 2017). Firstly, different solvents were used to prepare chitosan solutions such as TFA, acetic acid with concentrations in the range of 2%–90%. However, these did not work because of different problems relating to conductivity, viscosity, and surface tension. Chitosan was then blended with other spinnable copolymers such as PVA, PCL, and PLA.

Polyvinyl alcohol is one of the polymers previously blended with chitosan and tried for spinning, using the needle electrospinning method. PVA is soluble in water up to 9 wt%, which is a suitable concentration for spinning using the needle electrospinning process. So, in this research, 9 wt% PVA solution was tried for spinning, using the free surface wire electrospinning before blending with chitosan (Huang et al. 2007). The spinning was tried using 50 kv, and 75 kv, and carriage speed of 200 mm/sec.

Fig. 2 shows the SEM image of the fibers prepared from 9 wt% PVA solution. Obviously, the fibers started to form but at an extremely low rate. Therefore, the whole sample could be considered to mainly contain beads and lumps which indicate that during the process, the sample was merely spraying instead of spinning. Due to the inability of PVA to form fibers, it was not blended with chitosan.

Figure 2. 

Scanning electron microscope captures of 9 wt% PVA after spinning.

Polycaprolactone (PCL) is considered an ideal candidate that may be incorporated into chitosan, especially due to its salient features. Specifically, PCL is biocompatible, biodegradable, non-toxic and possesses sufficient mechanical properties (Schueren et al. 2011). Therefore, PCL is being investigated for biomedical applications. However, it is highly hydrophobic and does not have cell-recognition sites that can support cell adhesion (Prabhakaran et al. 2008). Interestingly, these properties can be supplied by chitosan. Hence, it is believed that incorporating PCL into chitosan would facilitate the electrospinning process of chitosan. Although PCL is not soluble in water, it is highly soluble in novel solvent such as acetic acid:formic acid mixture in the volume ratio of 3:7. The blend of PCL and chitosan was prepared by simultaneously dissolving them in a solvent mixture to get a final concentration of 10 wt% PCL with 2.5 wt% Cs (Schueren et al. 2012). Spinning was carried out using the Nanospider under voltage of 50 kv and carriage speed of 200 mm/sec. During the spinning process, fibers began to form and it was obvious to the naked eye as fibers were formed. However, as presented in Fig. 3, the SEM image of the prepared fibers of 10 wt% PCL with 2.5% Cs, revealed that the fibers formed are beaded.

It was thought that changing the solvent used may improve the spinnability or improve the quality of the nanofibers formed. Therefore, dichloromethane (DCM) was used to prepare a 3 wt% PCL solution blended with 2.5 wt% Cs, the spinning was carried out under voltage of 45 KV, carriage speed of 200 mm/sec, and 69% humidity. It should be noted that the spinning of organic solvents should be done at high humidity conditions to avoid evaporation. The SEM image of the prepared nanofibers is presented in Fig. 4. As can be seen from the figure, the formed fibers are beaded but much better than those obtained using the solvent mixture (Acetic Acid: Formic Acid, at 3:7). Hence, PCL cannot be fully considered as a suitable solvent to be blended with chitosan to improve its spinnability.

Figure 3. 

Scanning electron microscope captures of 10 wt% PCL blended with 2.5wt% Cs (Acetic Acid:Formic Acid) (3:7).

Figure 4. 

Scanning electron microscope captures of the nanofibers of 3 wt% PCL blended with 2.5 wt% Cs in DCM.

Poly lactic acid (PLA) is known to possess good spinning properties because of its surface activity and low viscosity which make it spinnable even at moderate to high concentrations. To determine the spinnable concentration of PLA, solutions with concentrations 2%, 4% and 6% in DCM were prepared and electrospun using a voltage of 45 Kv, and 200 mm/sec carriage speed. Morphologies of the obtained nanofibers are shown in Fig. 5. As can be seen from the figure, the concentrations 2% and 4% PLA formed beaded nanofibrous structure as presented in Fig. 5a and 5b). On the other hand, Fig. 5c shows the morphology of the 6% PLA electrospun nanofibers which exhibits a smooth, neat, and uniform surface. The average diameter is about 158.42±25.87 nm as obtained from the histogram of the fiber size distribution, estimated by the Image j software and shown in Fig. 8. The use of DCM as a solvent helped in getting good nanofibers, where DCM is considered low boiling point solvent (BP is 39.6 °C) compared with other solvents commonly used in electrospinning, such as dimethylformamide (DMF, 153 °C) and acetic acid (Ac.Ac. 118 °C). Solvents should be completely removed from the nanofibers, as the residual solvents may cause adverse effects when nanofibers are used in biomedical applications. Interestingly, this is one of the advantages of using DCM because it can be easily evaporated and then removed from the formed nanofibers (Ogawa et al. 2004).

Figure 5. 

Scanning electron microscope captures of (a) 2wt% PLA (b) 4%PLA (c) 6%PLA all in DCM as solvent.

PLA was then blended with Cs in DCM and tried for nanofibers preparation. Fig. 6 shows the prepared nanofibers of blended solution of 6 wt% PLA with 2 wt% 15 kDa Cs in DCM, under a voltage of 45 V, relative humidity of 69% and carriage speed of 200 mm/sec. The prepared nanofibers were uniformly distributed with an average diameter of about 200 nm. Notably, the presence of chitosan with PLA did not affect the quality of the nanofibers formed and their sizes were still around 200 nm.

Figure 6. 

Scanning electron microscope captures with measured sized fibers of the blends of 6wt% PLA with 2wt% Cs 15kDa nanofibers in DCM.

Generally, most of the previously reported work on chitosan nanofibers involved acidic solvents such as concentrated acetic acid (Sedghi et al. 2017), trifluoracetic acid (TFA) Haider et al. 2009), or a mixture of (TFA: DCM) in the ratio of 70:30 (Sangsanoh et al. 2006; Mendes et al. 2016). The main disadvantage of those nanofibers is that they could lose their structure when they come in contact with phosphate buffer saline (PBS, pH 7.4). This could arise from the dissolution of chitosan–acid salts that are usually formed when chitosan is dissolved in acids such as acetic acid or TFA (Sangsanoh et al. 2006). Therefore, the use DCM to dissolve chitosan herein will help to protect the structure of the produced nanofibers during application.

On the other hand, two concentrations of 15 kDa Cs were tried for the formation of nanofibers with 6% PLA. The images in Fig. 7 showed that the fibers formed from 15% 15 kDa Cs, with 6% PLA are considered beaded since large beads and lumps are obvious in the fibrous structure, while at higher concentration of chitosan (25% 15 kDa Cs, with 6% PLA) better beads-free nanofibrous structure was obtained.

The nanofibers diameter distribution diagrams of all prepared nanofibers are illustrated in Figs 8 and 9. Comparing the nanofiber diameters presented in Fig. 9 with the average nano fiber size of PLA (158.42 ± 25.87 nm), it can be inferred that the addition of chitosan to PLA did not cause wide changes in the nanofiber’s diameter. On the other hand, the results show that the influence of chitosan concentration on the nanofiber diameter was not significant in the case of 15 kDa Cs, where the size was almost the same in the two concentrations. This could be because solutions of the two different concentrations gave almost the same viscosity and surface tension.

Figure 7. 

Scanning electron microscope captures of (a) S1(15%Cs 15kDa) (b) S2(25%Cs 15kDa)

Figure 8. 

Histogram of the diameter distribution of 6%PLA with an average diameter size of 158.42 ± 25.87 nm.

Figure 9. 

Histogram of the diameter distribution of (a) S1(15% Cs 15kDa) and (b) S2 (25% Cs 15kDa).

The inter-molecular interaction between two blended polymers can be investigated through FTIR. The FTIR spectra of the different samples are illustrated in Fig. 10. As illustrated in the figure, the spectra of PLA exhibit some important peaks such as the band around 2999–2850 cm^-1 which is attributed to C-H stretching (Olabode et al. 2015). On the other hand, the peak at 1750 cm^-1 represents the stretching vibration of C=O groups from the carboxylic acid and ester groups of PLA. The peak at 1455 cm^-1 represents the characteristic stretching of –CH₃, whereas the C-H deformational peak can be seen at 1375 cm^-1 (Akindoyo et al. 2015). The major peaks available in the structure of chitosan include the peak around 3424 cm^-1 which represents the N-H stretching. On the other hand, the peak representing the bending of the NH₂ group appears around 1624–1629 cm^-1.

Notably, all peaks assigned to the saccharide structures appeared, where the peak around 2890 cm^-1 represents the C-H stretching peak, 1150 cm^-1 represents the C-N stretch peak, while the peak at 1320 cm^-1 is assigned to C-O stretching (Wang et al. 2006; Athamneh et al. 2013).

It is noteworthy that some distinct changes occurred in the FTIR spectra of chitosan-PLA nanofibers as illustrated in Fig. 10. Specifically, the FTIR spectra of all prepared grades of chitosan-PLA nanofibers revealed reduced intensity of the C=O stretching peak at 1750 cm^-1, as well as the reduction in the intensity of N-H stretching peak of chitosan. This is indication of interactions between the NH₂ groups of chitosan molecule and the carboxylic acid functional groups of PLA. Significantly, the reduction of the C=O peak was less when using higher concentration of chitosan (25%). This suggests that the reaction between chitosan and PLA was not complete. This low reduction in C=O peak intensity could be due to possible change in the conformation of chitosan, into a helical structure as the concentration increases. As such, more NH₂ groups will be hidden within the chains such that only a smaller number of NH₂ are available to interact with PLA. This would result in decreased interaction between PLA and chitosan. Table 2 summarizes the major peaks in the FTIR spectra of PLA, 15 kDa Cs, and Cs-PLA nanofibers.

Figure 10. 

FTIR spectra of PLA, Cs 15kDa, S1(15%Cs 15kDa) and S2(25%Cs 15kDa).

The DSC analysis helps to investigate the influence of chitosan molecular weight and concentration on the glass transition, crystallization, and melting phenomena of chitosan-PLA nanofibers. The DSC thermograms of the chitosan-PLA nanofibers are illustrated in Fig. 11. Obviously, the figure reveals distinct successive transition regions. The endothermic transitions represent the glass transition temperature (T_g), and the melting temperature (T_m). On the other hand, the exothermic transition represents the crystallization temperature, (Tc). Close observation of the curves shows that there is a slight right shift in the T_g of PLA (64 °C), following the incorporation of chitosan in all grades of the blended chitosan-PLA nanofibers. Chitosan is well known to have a rigid structure, with glass T_g of around 203 °C (Sakurai et al. 2000). Therefore, the right shift in T_g of PLA in the blends indicate partial miscibility between PLA and chitosan (Shuai et al. 2001; Suyatma et al. 2004). In another vein, around the Tg of polymers, the molecular chains often gain more flexibility and mobility (Jaszkiewicz et al. 2014). Therefore, the shift in the glass transition temperature indicates delay in the free mobility of PLA chains, which can be attributed to restrictions imposed by chitosan on the molecular chain mobility of PLA. This might have retarded the cooperative motions of its chains, thereby inducing an increase in the T_g (Liu et al. 2013).

Generally, the T_g values may be affected by the size of the side groups and the chain mobility. However, it is interesting to know that the T_g obtained herein are comparable to what was previously reported in previous studies (Cai et al. 2010; Au et al. 2012; Pachon et al. 2014). On the other hand, the T_m values show that the endothermic heat in the thermogram of chitosan-PLA nanofibers decreased after incorporating chitosan into the blends. Chitosan is an amorphous polymer and might have interfered with the formation of crystals in the structure of PLA in the chitosan-PLA blend. Hence, multi peaks can be seen in the melting peaks of chitosan-PLA nanofibers. Notably, the multiple peaks in the curve of PLA indicate that there is a different crystalline structure of PLA in the blends after electrospinning, which agrees with what was reported in a previous study (Au et al. 2012).

On the other hand, rigid chitosan in the 15% 15 kDa Cs, and 25% 15 kDa Cs could affect the formation crystalline structure of PLA during nucleation and crystallization. Therefore, this observation can be attributed to the amorphous nature of chitosan which might have retarded the rate of crystallization, thereby generating imperfect crystals (Sakurai et al. 2000). Summary of the DSC parameters such as T_g, and T_m of PLA and Chitosan-PLA blended nanofibers are summarized in Table 3.

Figure 11. 

DSC thermograms of Chitosan, PLA, S1 (15%Cs 15kDa), and S2 (25% Cs 15kDa).