The electrospinning behavior of poly(vinyl alcohol) in DMSO–water binary solvent mixtures

Abstract

Poly(vinyl alcohol) (PVA) is readily soluble in both water and DMSO, however, its solubility in a mixture of DMSO–water is highly dependent on the solvent composition. The present study reports on the electrospinning behavior of PVA solutions in DMSO–water binary mixtures at a temperature of 35 °C. Because of the strong interaction of water with DMSO, both the rheology and spinnability of the PVA solutions were influenced by the composition of the binary solvent mixture. At the same concentration of PVA, the morphology of the electrospun fibers changed from highly uniform thick fibers to very fine fibers with a large number of beads or droplets.

---

Introduction

Electrospinning has become the technology of choice for making nanofibers suitable for various applications like drug delivery, nanosensors, catalysts, filtration, etc.^1–3 Electrospinning can be easily manipulated by changing process parameters, ambient conditions and solution properties.^4–6 From our previous studies it has been established that the fiber diameter is majorly dependent on solution properties.⁷ The nature of solvent, which affects the polymer–solvent interaction, is an important aspect that alters the polymer solution properties and tends to influence its spinnability.^8,9

Solution properties, such as evaporation rate, surface tension or conductivity, have been modified by using multi-component solvent systems. Cellulose acetate, poly(ethersulfone), poly(caprolactone) and poly(styrene) have been electrospun using binary or ternary solvent systems.^10–14 Mixing of a good and a poor solvent has been used to produce porous nanofibers through non-solvent induced phase separation of PAN and PLA and improved morphology of gelatin nanofibers through changes in viscosity.^15–17

Poly(vinyl alcohol), a biocompatible and biodegradable polymer, is readily soluble in polar solvents like water and dimethyl sulfoxide (DMSO). Out of the two solvents, DMSO is reported to be a better solvent for PVA than water. Intrinsic viscosity [η] of PVA in DMSO is observed to be higher (3.25 dL g⁻¹) than in water (0.93 dL g⁻¹).¹⁸ High molecular weight of DMSO and high degree of its interaction with PVA, induce greater excluded volume that causes PVA chains to remain in extended conformation compared to that in water. Young et al.¹⁹ have shown using Flory–Huggins theory that the interaction parameter of water/PVA is 0.509 while that of DMSO/PVA is 0.328. Interestingly, when these two good solvents, DMSO and water, are mixed together, they are reported to behave as co-nonsolvents for PVA.²⁰ The two solvents show higher interaction between them (solvent–solvent) than their interaction with polymer (solvent–polymer). Solvent–solvent interaction peaks for 0.33 or 0.25 mole fraction of DMSO, where DMSO forms complex with either two or three molecules of water.²¹ Therefore, the solution rheology of PVA greatly varies with the composition of water and DMSO mixture. Gelation behavior and phase separation of PVA in DMSO–water mixture has been extensively studied by Takahashi et al. and Kanaya et al.^20,22 The solubility behavior of PVA in DMSO–water has been used for the production of high performance fibers using gel spinning.²³ The presence of DMSO in the system also allows better dispersibility of additives such as CNTs, which has been used to produce reinforced high performance PVA fibers.²⁴ This system may further be extended for creating blends of PVA with other polymers that are soluble in DMSO or for inducing phase separation and producing porous nanofibers by removing water soluble PVA. Additionally, DMSO and its mixtures with water find numerous applications in areas such as analytical chemistry, biology and medicine.²⁵ DMSO has very low toxicity that makes it suitable as pharmaceutical agent apart from being an effective solvent.²⁵ Therefore, the use of DMSO–water mixtures for dissolution of PVA can find many applications.

To the best of our knowledge, electrospinning of PVA and its blends has been reported only in single solvent systems.^26–29 Since PVA is known to exhibit peculiar rheological behavior in DMSO–water mixture, it is expected that composition of solvent mixture would have significant effect on the morphology of electrospun nanofibers. The aim of this study is to investigate the influence of DMSO–water composition on electrospinning behavior of PVA.

Experimental

Materials

High molecular weight PVA of M_w 125 000 (degree of hydrolysis (DH) ∼ 86–89%) was purchased from Central Drug House, New Delhi, India. De-ionized water (DI) and dimethyl sulfoxide (Merck, India) were used as solvents.

Preparation of solutions

10 wt% PVA in DMSO and water mixture with different weight ratios of DMSO and water were prepared (Table 1). Weighed amount of PVA was first dissolved in required amount of DMSO in a sealed bottle for each ratio and stirred at 90 °C in a water bath. The solutions were brought down to room temperature slowly and then required amount of DI water was added to make up the solvent ratios. The solutions were then stirred again to properly mix water with the remaining solution at temperature maintained at 35 °C. Freshly prepared solutions were used for electrospinning and rheological studies to avoid gelling.

Table 1 Rheological properties and spinning behavior of 10 wt% PVA solutions in different ratios of DMSO–water mixtures

Sample name	Wt of DMSO (%)	χDMSO (mole fraction)	Viscosity (Pa s)@100 s^−1	Storage modulus (Pa)@100 rad s^−1	Relaxation time (ms)	Fiber diameter (nm)	Spinning behavior
0 DMSO	0	0.0	0.752	9.69	0.682	236 ± 20	Uniform fibers
20 DMSO	20	0.05	1.127	15.02	1.002	558 ± 79	Uniform fibers
30 DMSO	30	0.09	1.483	22.17	1.345	571 ± 42	Uniform fibers
40 DMSO	40	0.13	1.908	34.26	1.451	603 ± 205	Non-uniform fibers
50 DMSO	50	0.19	2.55	50.15	1.924	456 ± 157	Non-uniform fibers
60 DMSO	60	0.26	2.89	62.59	2.203	112 ± 28	Fine fibers + beads
70 DMSO	70	0.35	2.711	55.92	2.019	94 ± 19	Fine fibers + beads
80 DMSO	80	0.48	2.82	52.93	1.841	87 ± 17	Fine fibers + droplets
100 DMSO	100	1.0	2.084	33.36	1.621	68 ± 13	Fine fibers + droplets

---

Rheological characterization of PVA–DMSO–water solutions

The PVA–DMSO–water solutions were characterized using rotational rheometer (MCR 302, Anton Paar GmBh, Germany) for their oscillatory and shear rheological measurements with 50 mm diameter and 1° cone-plate geometry. All the polymer solutions of PVA–DMSO–water were thermally equilibrated at 35 °C before carrying out the test. The strain amplitude (γ°) as obtained from amplitude sweep was 10%. The steady state measurements were conducted at shear rates from 0.01 to 1000 s⁻¹ at 35 °C. Frequency sweeps were conducted from 0.1 to 100 rad s⁻¹ at 35 °C.

Interfacial rheological tests were carried out using interfacial cell of the same rheometer with bicone geometry of diameter 68.28 mm, cone angle 5° and penetration depth of 2.213 mm. The volume of all the solutions tested was approximately 114 mL each. Interfacial viscosity measurements were made at shear rate ranging from 0.1 to 100 s⁻¹. The values of interfacial viscosity of different samples were compared at shear rate of 0.1 s⁻¹.

Electrospinning of PVA solutions

A bipolar system with high voltage power supply (D-ES40PN-10W, Gamma High Voltage Research, Ormond Beach, FL, USA) was used for electrospinning. The polymer droplet was charged conductively by connecting spinning needle with a positive voltage and inductively by connecting collector with a negative voltage. Polymer solutions were extruded using a syringe pump (KDS-100-CE, KD Scientific, Holliston, MA, USA), 2 mL syringe and blunted needle of 18 G. Fibers were deposited on circular aluminum plate used as collector. Nanofiber samples were collected at flow rates of 0.5 mL h⁻¹ and at needle to collector distance of 20 cm. PVA–DMSO–water solutions thermally equilibrated at 35 °C were electrospun in a chamber maintained at 35 °C and relative humidity of 40% to ensure efficient evaporation and drying of the solvents. The electrospun fibers, thus obtained, were vacuum dried for overnight before their characterization.

Surface morphology of electrospun nanofibers

Surface morphology of the fibers was obtained using Scanning electron microscope (Quanta 200F, FEI, Netherlands). The fiber diameters were determined using Image J software. For each sample an average diameter of two hundred fibers with standard deviation has been reported.

Results and discussion

Solution properties

PVA has a natural tendency to form gels in various solvents when quenched at lower temperatures. The gelling occurs due to the formation of crosslink points or crystallites. Since DMSO is a better solvent for PVA, solutions of PVA–DMSO are more transparent than that of PVA–water.³⁰ For DMSO–water mixtures it has been reported in the literature that PVA–DMSO–water form gels when quenched at lower temperatures. The transparency of the gels has been reported to be dependent on the quenching temperature and DMSO–water ratio. It has been reported by Hyon et al. that at DMSO to water ratio of 0.8 (volume ratio) transparent gels were formed but at DMSO to water ratio of 0.6 (volume ratio) opaque gel were formed that eventually turned transparent when quenched at temperatures less than 20 °C.³¹ They also reported that critical concentration for gelling increases with increase in quenching temperature which imply that at higher temperatures gelling takes longer time. In the present study, all PVA–DMSO–water solutions were prepared at 90 °C and slowly brought down to 25 °C. A visual assessment showed that the solutions were transparent suggesting the formation of homogenous solutions (Fig. 1).

Fig. 1 Photograph of 10 wt% PVA–DMSO–water solutions showing optical transparency.

Fig. 2 shows flow curves of 10 wt% PVA in DMSO–water mixtures. Viscosity increases steadily to 2.89 Pa s with increase in the DMSO till DMSO to water ratio of 0.6. On further increasing the DMSO content to 70 and 80%, a slight decrease in the viscosity is observed to 2.71 and 2.82 Pa s, respectively, with a variation of 0.09 Pa s only. Viscosity of PVA in 100% DMSO (100 DMSO) show a much lower value of 2.08 Pa s, however, it is still higher than the viscosity in 100% water (0 DMSO). The viscosity of solvent mixture is a function of their composition, which in turn affects the viscosity of PVA solutions in these mixtures. At 70 DMSO, the viscosity of the solvent mixture itself is higher than that of 100 DMSO, therefore, resulting in higher viscosity values. The viscosity of PVA solution in 100% DMSO is higher than that in 100% water because the PVA chains tend to acquire an extended conformation in DMSO exhibiting higher interaction among the chains through DMSO molecules. Solutions with higher water content show Newtonian behavior up to 100 s⁻¹ thereafter shear-thinning is seen. However, solutions with higher DMSO content show shear-thinning behavior after 50 s⁻¹ shear rate.

Fig. 2 Viscosity vs. shear rate curve of 10 wt% PVA in different DMSO–water ratios.

Frequency sweep curve also shows similar trend as that of flow curve. The storage (G′) and loss moduli (G′′) of PVA solutions increase up to DMSO–water mixture of 60 : 40. On further increasing DMSO content, the storage and loss moduli start to decrease (Fig. 3a and b). Slope of G′ between angular frequency (ω) of 0.1 and 1 rad s⁻¹ is 0.11 and shows a plateau like region. This can be attributed to the presence of intermolecular hydrogen bonding and is a characteristic of PVA solutions. On further increase of ω in the range of 1 to 500 rad s⁻¹ the slope abruptly increases to 1.56 whereas the slope of G′′ remains nearly 1.

Fig. 3 (a) Loss and (b) storage modulus of 10 wt% PVA in various DMSO–water mixtures.

Logarithmic plot of storage modulus vs. loss modulus, known as Cole–Cole plot, is shown in Fig. 4a. The slope of 2 signifies the homogeneity of a solution. However, the slope in modified Cole–Cole plot of the PVA solutions in DMSO–water mixture is nearly 1.77, which may be due to stronger intermolecular interactions among PVA chains.³² The slope value of less than 2 suggests that the solutions are heterogeneous in spite of being optically transparent.

Fig. 4 (a) Modified Cole–Cole plot and (b) plot of loss tangent vs. angular frequency for 10 wt% PVA–DMSO–water solutions.

Plot of loss tangent (tan(δ)), which is a ratio of G′′ to G′, is a measure of liquid-like viscous behavior signified by G′′ and solid like elastic behavior signified by G’. The equilibrium point of G′ and G′′ (i.e. tan(δ) = 1) indicates gelling point. From the graph (Fig. 4b) it is evident that value of tan(δ) first increases up to angular frequency of 5 rad s⁻¹, which implies that solutions are becoming more liquid-like possibly due to breakage of strong inter-molecular H-bonding. On further increase of angular frequency (beyond 10 rad s⁻¹) there is a decrease in tan(δ) which indicates increase in solid-like (i.e. elastic) behavior, which may be due to improved association of unfolded chains at higher shear. However, sol–gel transition was not seen in any of the solutions. This implies that in-homogeneity seen from the modified Cole–Cole plot is not due to liquid–solid type phase separation but more of liquid–liquid type phase separation into polymer rich and polymer lean phases.

Of all the compositions, PVA solution in DMSO : water ratio of 60 : 40 has the lowest value of tan(δ) at high frequencies (in the range of 5–500 rad s⁻¹), which indicates that the solution has the highest elasticity.

Fig. 5a shows the viscosity and relaxation time (indicative of elasticity) of the PVA solutions in DMSO–water mixtures at different solvent ratios. Clearly both the viscosity and the relaxation time increase with increasing DMSO content till a DMSO–water ratio of 60 : 40. Thereafter, there is a small decrease in the two parameters.

Fig. 5 (a) Plot of viscosity and relaxation time of PVA–DMSO–water solutions vs. mole fraction of DMSO. (b) Interfacial viscosity vs. mole fraction of DMSO for PVA–DMSO–water solutions.

Interfacial rheology has been recently^29,33 shown to be important in predicting electrospinnability of polymers particularly because in electrospinning there is a significant increase in the surface to volume ratio as the jet thins down. Interfacial rheology has been considered as complimentary to the bulk rheology and has been used by Rosic et al.³⁴ to better correlate nanofiber morphology and electrospinnability of chitosan and alginate blended with PEO. In the present study, interfacial rheology was performed on the PVA–DMSO–water solution with different ratios of DMSO and water. The interfacial viscosity at a shear rate of 0.1 s⁻¹ was plotted against the mole fraction of DMSO as shown in Fig. 5b. As obvious from the figure, the interfacial viscosity follows a similar trend as seen in the case of bulk viscosity shown in Fig. 5a. In both cases the maximum value of viscosities were observed at 60 DMSO content. This indicates that the nature of interactions between the solvents and polymer remain similar in both interfacial and bulk phases of the spinning systems.

These solutions were electrospun to understand the effect of solvent composition on the spinning behavior of various solutions.

Electrospinning of PVA–DMSO–water solutions

Electrospinning process is initiated as the polymeric jet emerges from Taylor cone. This jet can be divided into two sections – first the straight part just before the whipping region caused due to the ability of the polymer to resist the breakup, and second, the instability region caused by electrostatic repulsion. In this region, the jet may split into many fibers and undertakes helical path due to bending instability.³⁵ The jet length depends on the rheology of the solution and the extent of electrostatic forces experienced by the spinning jet. It has been reported by Yu et al. that the length of the straight segment of the jet increases with increase in elasticity.³⁶

Interestingly, when the solution of PVA in DMSO–water mixtures were electrospun, their spinning behavior was markedly different. They formed different jet-lengths as depicted in Fig. 6. The jet length was observed to increase with increase in relaxation time of the polymer solutions till a DMSO–water ratio of 40 : 60. The jet length started to decrease on further increasing the DMSO content though the elasticity of the solution was higher at DMSO–water ratios of 50 : 50 and 60 : 40.

Fig. 6 Photographs of electrospinning jet lengths (straight path of jet after Taylor cone) of 10 wt% PVA solutions in various DMSO–water ratios. Solution with 40 wt% DMSO (40 DMSO) shows maximum jet length.

Fig. 7a–i shows SEM micrographs of PVA nanofibers electrospun from solutions in different DMSO–water ratios. PVA solution (10 wt%) in water could result in uniform fibers of fine diameter of 236 ± 20 nm. As the DMSO content in the solvent mixture was increased to 20 and 30 wt% (mole fraction 0.05, 0.09), the fibers remained uniform, however, the diameter increased significantly to 558 ± 79 and 571 ± 42 nm respectively. On further increasing the DMSO to 40 wt% (mole fraction 0.13), the fibers started to become highly non-uniform resulting in high variation in diameter (603 ± 205 nm). When PVA was electrospun using solvent mixture having 50 wt% DMSO, the diameter was observed to decrease while the non-uniformity remained high as before (456 ± 157 nm). With DMSO content of 60 and 70 wt%, PVA solutions were converted to very fine nanofibers with diameters of 94 ± 19 and 87 ± 17 nm respectively along with large number of beads. At 80 and 100 wt% DMSO, the conversion of polymer solution to nanofibers was found to be incomplete. The morphology as observed under SEM showed nanofibers with very fine diameter of 68 ± 13 nm covered with a large number of polymer droplets. The results with changing solvent composition are summarized in Fig. 8. The effective voltage required to electrospin the PVA solutions changed with changing DMSO–water composition. PVA solutions with 0 wt% DMSO was spun at 14 kV. With increase in DMSO content from 20 to 40 wt%, the required voltage decreased to 12.5 kV for 20 DMSO, 11 kV for 30 DMSO and 8 kV for 40 DMSO. On further increasing the DMSO content, the required voltage increased to 10 kV for 50 DMSO. For DMSO content of 60–100 wt%, the electrospinning voltage increased gradually from 15.5 kV for 60 DMSO to 17 kV for 100 DMSO.

Fig. 7 SEM images of electrospun nanofibers of 10 wt% PVA at different DMSO–water ratios showing four morphological regions, (a)–(c) show uniform fibers, (d) and (e) show non-uniform fibers, (f) and (g) show fine fibers with beads, (h) and (i) show finer fibers with droplets.

Fig. 8 Graph of diameter vs. DMSO mole fraction with schematic representation of polymer–solvent interaction at various mole fraction of DMSO. Region (I) of the graph shows uniform fibers due to free water molecules. Regions (II) and (III) show the liquid–liquid demixing that leads to non-uniform fibers and beaded fibers. Region (IV) shows fibers with droplets due to free DMSO molecules.

At DMSO content of 60 and 70 wt%, the DMSO to water molar ratio is close to 1 : 3 and 1 : 2, respectively. At these molar ratios, solution properties of PVA/DMSO/water ternary system are known to be greatly influenced by DMSO–water interaction.¹⁹ DMSO and water are observed to form DMSO/(water)₂ or DMSO/(water)₃ complexes at all ratios of DMSO–water mixtures. At lower DMSO mole fraction, all the DMSO molecules are engaged in forming complex with water leaving a few water molecules to interact with PVA molecules.

On the other hand, at higher DMSO mole fraction, all the water molecules form complex with DMSO molecules and a few DMSO molecules are left to interact with PVA molecules. At DMSO mole fractions of ∼0.33 and ∼0.28, the DMSO and water have the maximum association with each other leaving the PVA molecules to form a highly contracted coiled chain conformation.²¹ It was concluded by Takahashi et al. that DMSO with mole fraction of 0.33 is the poorest solvent for PVA.²⁰ In our case 60 DMSO corresponds nearly to mole fraction of ∼0.26 and 70 DMSO as ∼0.35. However, the temperature used in our study is higher at 35 °C compared to the above mentioned studies, which is likely to create a dynamic equilibrium between solvent–solvent and solvent–polymer interaction. Therefore, 60 and 70 wt% DMSO remain a solvent for PVA, though with poor solvency, and does not undergo solid–liquid type phase separation (Fig. 8).

The electrospinning behavior may be explained based on the above interaction of the two solvents. As the DMSO mole fraction is increased in the solvent mixture, the PVA is likely to get entangled leading to an increase in both viscosity and relaxation time resulting in higher diameter of the electrospun fibers. Interfacial viscosity, which follows the same trend as bulk rheology, also supports this observation. However, the uniform fibers are formed till sufficient water is available as a solvent for PVA. At higher mole fraction of DMSO, there appears to be liquid–liquid demixing resulting in still higher diameter of the nanofibers but with high variability. At the DMSO–water ratios of 60 : 40 and 70 : 30, which correspond to 1 : 3 and 1 : 2 mole ratios of DMSO–water, the polymer chains are likely to be phase separated in polymer rich and polymer lean phases.

The polymer rich phase is likely to give either non-uniform thick fibers or large beads, whereas the polymer-lean phase is likely to result in the formation of fine fibers or fine droplets in cases where concentration of PVA in lean phase is below the entanglement regime (Fig. 8). Polymer rich- and lean phases have been shown to create variable morphology in electrospun poly(vinylidene fluoride) nanofibers, where solvent rich phase dispersed in a matrix of polymer rich phase led to the formation of pores on fibers.³⁷

Further increase in DMSO content, where DMSO is in excess, allows better solubilization of PVA. As already explained earlier, in presence of DMSO, PVA molecules acquire an extended conformation because of which there is more intermolecular H-bonding and less entanglements as compared to that in water. The presence of low entanglements due to the extended PVA chains results in formation of droplets along with the very fine fibers. Similar behavior is also observed when pure DMSO is used as a solvent. The entanglements among the chains may be improved by increasing the concentration of PVA above 10 wt%, which may facilitate the formation of fibers with higher diameter without droplets. Fig. 9 shows defect free electrospun nanofibers can be obtained at a higher concentration of 12 and 14 wt% PVA in pure DMSO.

Fig. 9 SEM images of (a) 10 wt% PVA (b) 12 wt% PVA and (c) 14 wt% PVA in pure DMSO. Increase in the PVA concentration in DMSO leads to defect free fibers at 12 and 14 wt% PVA.

From the above results, it may be inferred that relationship between morphology and rheology can be explained only to a DMSO content of 40 wt%. Thereafter, this relationship is not evident possibly due to inhomogeneity of the solutions with higher DMSO content in the range of 50–70 wt%. The study has emphasized the role of liquid–liquid demixing at molecular level on fiber morphology, which in turn depends upon solvent–solvent and solvent–polymer interactions at various DMSO to water ratios. Interestingly, interaction between two good solvents, which can result in the phase separation, can significantly influence the morphology of the electrospun fibers.

Conclusions

The solutions of PVA in mixtures of DMSO and water were studied to understand the effect of interaction between the two good solvents on the rheology and spinning behavior of PVA. Rheological properties of PVA solutions varied significantly with different DMSO–water molar ratios at the same concentration of PVA. Since DMSO and water show high affinity towards each other, resultant viscosity (bulk and interfacial) and relaxation time were found to be very high at DMSO : water molar ratios of 1 : 2 & 1 : 3. The high values of viscosities and relaxation time were attributed to liquid–liquid demixing at these ratios, which were indicated by Cole–Cole and tan(δ) plots. Consequently, the morphology of the electrospun nanofibers obtained from these solutions changed with DMSO : water ratio. At lower DMSO content (i.e. high free water content), uniform fibers were formed and the diameter increased from 236 ± 20 for 0 DMSO to 603 ± 205 for 40 DMSO solutions, showing a good correlation with the relaxation time till 40% DMSO content. As the DMSO : water molar ratios approached 1 : 2–1 : 3 (i.e. 50, 60 and 70 wt% DMSO), the solutions resulted in poor spinnability with formation of non-uniform fibers with droplets due to prevailing inhomogeneity at these conditions. This liquid–liquid demixing is expected to cause polymer rich- and lean-phases, which result in formation of two different morphologies on electrospinning. At very high DMSO content, the spinning behavior was similar to that observed with PVA–DMSO system. The fiber diameters obtained from PVA–DMSO–water ternary system is a result of the interplay between the solvent–solvent and solvent–polymer interactions at various DMSO to water ratios.

Acknowledgements

We acknowledge partial financial support provided by the Department of Science and Technology, India under various research grants.

Notes and references

1. H. Cho, S. Y. Min and T. W. Lee, Macromol. Mater. Eng., 2013, 298, 475–486.
2. H. Zhang and X. Qiam, Adv. Mater. Res., 2011, 148–149, 1138–1143.
3. F. Dotti, A. Varesano, A. Montarsolo, A. Aluigi, C. Tonin and G. Mazzuchetti, J. Ind. Text., 2007, 37, 151–162.
4. T. Subbiah, G. S. Bhat, R. W. Tock, S. Parameswaran and S. S. Ramkumar, J. Appl. Polym. Sci., 2005, 96, 557–569.
5. S. Ramakrishna, K. Fujihara, W. Teo, T. Lim and Z. Ma, An introduction to electrospinning and nanofibers, World Scientific Publishing Co. Pvt. Ltd., Singapore, 1st edn, 2005.
6. W. E. Teo and S. Ramakrishna, Nanotechnology, 2006, 17, 89–106.
7. S. Basu, M. Jassal and A. K. Agrawal, Fibres Polym., 2013, 14, 950–956.
8. V. E. Dreval, A. Y. Malkin and G. V. Vinogradov, Eur. Polym. J., 1973, 9, 85–99.
9. Y. Isono and M. Nagasawa, Macromolecules, 1980, 13, 862–867.
10. S. Tungprapa, T. Puangparn, M. Weerasombut, I. Jangchud, P. Fakum, S. Semongkhol, C. Meechaisue and P. Supaphol, Cellulose, 2007, 14, 563–575.
11. G. Darko, A. Goethals, N. Torto and K. De Clerck, Appl. Nanosci., 2015, 1, 1–9.
12. C. J. Luo, E. Stride and M. Edirisinghe, Macromolecules, 2012, 45, 4669–4680.
13. T. Uyar and F. Besenbacher, Polymer, 2008, 49, 5336–5343.
14. G. Eda, J. Liu and S. Shivkumar, Eur. Polym. J., 2007, 43, 1154–1167.
15. K. Nayani, H. Katepalli, C. S. Sharma, A. Sharma, S. Patil and R. Venkataraghavan, Ind. Eng. Chem. Res., 2012, 51, 1761–1766.
16. J. Oliveira, J. Eng. Fibers Fabr., 2014, 9, 118–125.
17. N. Choktaweesap, K. Arayanarakul, D. Aht-ong, C. Meechaisue and P. Supaphol, Polym. J., 2007, 39, 622–631.
18. S. J. Hong, H. T. Huang and P. D. Hong, J. Appl. Polym. Sci., 2004, 92, 3211–3217.
19. T. H. Young and W. Y. Chuang, J. Membr. Sci., 2002, 210, 349–359.
20. N. Takahashi, T. Kanaya, K. Nishida and K. Kaji, Polymer, 2003, 44, 4075–4078.
21. S. J. Hong, P. D. Hong, J. C. Chen and K. S. Shih, Eur. Polym. J., 2009, 45, 1158–1168.
22. T. Kanaya, M. Ohkura, K. Kaji, M. Furusaka and M. Misawa, Macromolecules, 1994, 27, 5609–5615.
23. W. I. Cha, S. H. Hyon and Y. Ikada, J. Polym. Sci., Part B: Polym. Phys., 1994, 32, 297–304.
24. P. Zhang, T. Zhou, L. He, S. Zhang, J. Sun, J. Wang, C. Qin and L. Dai, RSC Adv., 2015, 5, 55492–55498.
25. J. Catalan, C. Diaz and F. Garcia-Blanco, J. Org. Chem., 2001, 66, 5846–5852.
26. P. Supaphol and S. Chuangchote, J. Appl. Polym. Sci., 2008, 108, 969–978.
27. J. C. Park, T. Ito, K. O. Kim, K. W. Kim, B. S. Kim, M. S. Khil, H. Y. Kim and I. S. Kim, Polym. J., 2010, 42, 273–276.
28. M. Shahi, A. Moghimi, B. Naderizadeh and B. Maddah, Sci. Iran., 2011, 18, 1327–1331.
29. R. Rosic, J. Pelipenko, J. Kristl, P. Kocbek, M. Bester-Rogac and S. Baumgartner, Eur. Polym. J., 2013, 49, 290–298.
30. Y. Hou, C. Chen, K. Liu, Y. Tu, L. Zhang and Y. Li, RSC Adv., 2015, 5, 24023–24030.
31. S. H. Hyon, W. I. Cha, Y. Ikada, M. Kita, Y. Ogura and Y. Honda, J. Biomater. Sci., Polym. Ed., 1994, 5, 397–406.
32. Y. H. Cho, K. S. Dan and B. Chul, Korea Aust. Rheol. J., 2008, 20, 73–77.
33. O. Regev, S. Vandebril, E. Zussman and C. Clasen, Polymer, 2010, 51, 2611–2620.
34. R. Rosic, J. Pelipenko, P. Kocbek, S. Baumgartner, M. Bester-Rogac and J. Kristl, Eur. Polym. J., 2012, 48, 1374–1384.
35. D. H. Reneker, A. L. Yarin, H. Fong and S. Koombhongse, J. Appl. Phys., 2000, 87, 4531–4547.
36. J. H. Yu, S. V. Fridrikh and G. C. Rutledge, Polymer, 2006, 47, 4789–4797.
37. H. Fashandi, A. Yegane and M. M. Abolhasani, Fibres Polym., 2015, 16, 326–344.

---