International Journal of Biological Macromolecules 80 (2015) 139–148

Contents lists available at ScienceDirect

International Journal of Biological Macromolecules journal homepage: www.elsevier.com/locate/ijbiomac

Improving agar electrospinnability with choline-based deep eutectic solvents Ana M.M. Sousa a,b , Hiléia K.S. Souza a , Joseph Uknalis b , Shih-Chuan Liu b,c , Maria P. Gonc¸alves a , LinShu Liu b,∗ a

REQUIMTE/LAQV, Faculdade de Engenharia, Universidade do Porto, Rua Dr. Roberto Frias, 4200-465 Porto, Portugal Dairy Functional Food Research Unit, United States Department of Agriculture, Agricultural Research Service, Eastern Regional Research Center, 600 East Mermaid Lane, Wyndmoor, PA 19038, USA c School of Health Diet and Industry Management, Chung Shan Medical University; Hospital Department of Nutrition, Chung Shan Medical University. 110, Sec. 1, Jian-Guo North Rd., Taichung, 402, Taiwan, ROC b

a r t i c l e

i n f o

Article history: Received 18 March 2015 Received in revised form 12 June 2015 Accepted 18 June 2015 Available online 23 June 2015 Keywords: Agar Electrospinning Microfibers Nanofibers Deep eutectic solvents PVA

a b s t r a c t Very recently our group has produced novel agar-based fibers by an electrospinning technique using water as solvent and polyvinyl alcohol (PVA) as co-blending polymer. Here, we tested the deep eutectic solvent (DES), (2-hydroxyethyl)trimethylammonium chloride/urea prepared at 1:2 molar ratio, as an alternative solvent medium for agar electrospinning. The electrospun materials were collected with an ethanol bath adapted to a previous electrospinning set-up. One weight percent agar-in-DES showed improved viscoelasticity and hence, spinnability, when compared to 1 wt% agar-in-water and pure agar nanofibers were successfully electrospun if working above the temperature of sol–gel transition (∼80 ◦ C). By changing the solvent medium we decreased the PVA concentration (5 wt% starting solution) and successfully produced composite fibers with high agar contents (50/50 agar/PVA). Best composite fibers were formed with the 50/50 and 30/70 agar/PVA solutions. These fibers were mechanically resistant, showed tailorable surface roughness and diverse size distributions, with most of the diameters falling in the sub-micron range. Both nano and micro forms of agar fibers (used separately or combined) may have potential for the design of new and highly functional agar-based materials. Published by Elsevier B.V.

1. Introduction In recent years, the electrospinning technique has become the preferred method for the fabrication of polymer nano and microfibers with high surface-to-mass ratios, enhanced mechanical performance and high porosity [1–4]. Briefly, a high voltage is applied to the polymer solution or melts held at a tip of a capillary by its surface tension. When the applied electric field is such that overcomes the surface tension of the droplet, a jet forms and stretches towards the grounded collector (or counter electrode). If the solvent is volatile, it will evaporate as the jet travels towards the counter electrode [5,6]. However, if non-volatile solvents are used (e.g. ionic liquids, ILs), the fibers should be recovered through the use of a coagulating bath [7,8]. Despite this apparent simplicity, an intricate conjugation of factors is involved in electrospinning. Parameters that govern the spinnability include

∗ Corresponding author. Tel.: +1 215 233 6486. E-mail address: [email protected] (L. Liu). http://dx.doi.org/10.1016/j.ijbiomac.2015.06.034 0141-8130/Published by Elsevier B.V.

the ambient temperature and relative humidity, the feed rate of the solution, the applied electric field, and the distance from the capillary tip-to-collector [2]. Other aspects related to the properties of the polymer solution (e.g. concentration, viscosity, viscoelasticity, surface tension) as well as the polymer and solvent intrinsic properties are also crucial for the electrospinning process. Micro and nano forms of polymers can be equally advantageous in the design of new products [1,2,9]. Several cutting-edge materials have also been developed from the combination of nanofibers with micro-scale structures. Collagen nanofibers and starch-based microfibers have been combined to design tissue scaffolds benefiting from the great biochemical functionality of the former and the superior mechanical strength and greater porosity of the latter [9]. Edwards et al. developed a composite scaffold formed when electrospinning poly(lactic-co-glycolic acid) nanofibers onto a micro-scale multiwalled carbon nanotube [10]. Lee and Kim combined electrospun fibers and microsized structures of polycaprolactone with cell-embedded alginate-based hydrogels to develop hierarchical scaffolds for hard tissue engineering [11].

140

A.M.M. Sousa et al. / International Journal of Biological Macromolecules 80 (2015) 139–148

Agar is a natural polymer extracted from selected red seaweeds mainly used as gel in food and biotechnological applications [12,13]. Properties such as biocompatibility, biodegradability and non-toxicity, make agar also ideal for more sensitive fields such as the biomedical [14–16]. Knowing that nano and microfibers may afford performances and properties not seen in bulk materials [2,9,17], processing agar down to these scales could be a major breakthrough in the development of cutting-edge and highly functional biomaterials. Very recently, our group has taken the first steps to fabricate agar fibers by an electrospinning technique using water as solvent and polyvinyl alcohol (PVA) as co-blending polymer [18]. Agar containing PVA nanofibers were successfully produced using a tubeless spinneret attached inside the electrospinning chamber, set to operate at 50 ◦ C. In this first approach however, the pure agar solution (i.e. 1 wt% agar-in-water) showed inadequate spinnability and only blends with high PVA contents (30/70 and 20/80 agar/PVA mass ratios prepared from a 10 wt% PVA starting solution) yielded continuous fibrous mats. The choice of solvent is a decisive element in electrospinning [2]. In many cases, ILs have shown to be good alternatives to volatile solvents (e.g. water) [7,8,19]. These solvents are highly versatile and extremely efficient at dissolving natural polymers and this has led to a growing interest in natural fibers made from ILs in fields such as the biomedical [20]. Deep eutectic solvents (DESs), also known as IL analogues, are much cheaper and easier to prepare than ILs and for that reason, are currently being focus of a lot of attention [21,22]. The DES concept arises from the adequate mixture of halide salts with hydrogen bond donors to form eutectics that can be operated at room temperature as fluids. Particular emphasis has been given to DESs derived from the quaternary ammonium salt (2-hydroxyethyl)trimethylammonium chloride (choline chloride, ChCl), due to its very low cost, biodegradability, non-toxicity as well as renewable nature [21,22]. Very recently, Mukesh et al. produced chitin nanofibers by sonication using a DES based on ChCl and thiourea [23]. A eutectic mixture with a freezing point of 12 ◦ C is also formed when mixing ChCl (melting point, m.p. = 302 ◦ C) and urea (m.p. = 133 ◦ C) at 1:2 molar ratio [21]. In the present study, we tested for the first time the DES ChCl/urea prepared at 1:2 molar ratio, as an alternative solvent medium to process agar by electrospinning. To the best of our knowledge, this is the first report on production of electrospun agar nanofibers without the use of a co-blending polymer as well as the first time this DES is used as solvent medium for electrospinning. Due to the non-volatile nature of the DES a coagulating ethanol bath was used to collect the fibers. The solvent bath was adapted to an electrospinning set-up used in a previous study where we had used a rotating cylindrical drum to collect agar/PVA nanofibers from aqueous solutions [18]. PVA was added to agar [24,25] to further improve agar-in-DES spinnability and produce composite fibers with tailorable morphology. The morphology of the fibers was examined by scanning electron microscopy (SEM) and interpreted in light of the rheological properties of the spinning solutions.

molecular mass, Mv ∼138 kDa; 3,6-anhydro-␣-L-galactose, 3,6-AG (%) ∼44; sulfate esters, SO3 − (%) ∼1.6. Details concerning the followed experimental protocols used for agar characterization can be found elsewhere [26]. The choline chloride (>98%; C5 H14 ClNO) and urea (>99%; CH4 N2 O) were also purchased from Sigma–Aldrich. 2.2. Methods 2.2.1. Preparation of the DES The choline-based DES was prepared according to a previous procedure [27]. Briefly, the dried ChCl (70 ◦ C overnight in an oven) was mixed with the appropriate amount of the H-bond donor, i.e., urea at 1:2 molar ratio. The mixture was heated at 70 ◦ C under stirring until formation of a homogeneous liquid. Subsequently, it was left to cool down to 25 ◦ C and kept at that temperature (above the freezing point of the solvent).

2.2.2. Preparation of agar/PVA spinning solutions Agar and PVA powders were pre-dried overnight at 40 ◦ C in a vacuum oven, prior to use. The following concentrations were considered for the starting solutions: 1 wt% (agar) and 5 wt% (PVA). Both concentrations were defined according to preliminary tests. Each starting solution was prepared separately considering a solution volume of 15 mL and taking into consideration the density of the solvent (1.25 g/cm3 at 25 ◦ C; [22]). The appropriate amount of agar (0.15 g) or PVA (0.75 g) was mixed with the DES and heated in an oil bath, in closed cap vials, with temperatures ranging between 120 and 130 ◦ C and under vigorous stirring. In each case, the solution was initially heated to 120 ◦ C, kept at this temperature for 20 min after which, the temperature of the oil bath was increased to 130 ◦ C until a homogeneous solution was obtained. This typically occurred after 20–30 min of total dissolution time. Blends with different mass ratios (100/0, 50/50, 30/70, 20/80, 0/100 agar/PVA) were prepared by weighing the appropriate amounts of each starting solution into closed cap vials. Finally, the mixtures were heated again at 120 ◦ C, under vigorous stirring, until a homogeneous solution was obtained (∼20 min).

2.2.3. Rheological measurements The rheological characterization of the spinning solutions was performed in a stress-controlled rheometer (ARG2, TA Instruments, USA) as described elsewhere [18]. Each sample was degassed in a vacuum oven for 5 min at 100 ◦ C before being placed on the hot Peltier (90 ◦ C). A cone-and-plate (4 cm diameter, 2◦ angle and a 54 ␮m gap) was the chosen geometry to carry out the experiments. Frequency scans were recorded after cooling the solutions from 90 to 50 ◦ C at a cooling rate of 1 ◦ C/min and a fixed angular frequency, ω, of 6.28 rad/s, followed by an equilibration period at 50 ◦ C. Mechanical spectra were recorded at 50 ◦ C, over the range 0.1–75 rad/s. The relationship between the magnitude of the complex viscosity, | * |, and ω, was determined using the power law equation (Eq. (1)) [28],

2. Materials and methods

| ∗ | = Kωn

2.1. Materials

where K is the dynamic consistency index and n the dynamic power-law factor. In the limit, n can assume the value −1 (completely elastic system) or 0 (completely viscous system). The strain conditions were chosen according to the linear viscoelastic region defined during strain sweep tests. A common strain of 2% was selected for all samples. Steady shear measurements were carried out at 50 ◦ C, in the range of shear rates 1–200 s−1 . Three replicates were performed in each case.

The PVA (average Mw = 89,000–98,000 Da, 99+% hydrolyzed) and commercial agar (A-7002, (C12 H18 O9 )n) were both purchased from Sigma–Aldrich Co. (St. Louis, MO) and were the same as used in our electrospinning study focusing agar aqueous solutions [18]. This commercial agar sample was previously characterized in our lab [26] following standard procedures: viscosity-average

(1)

A.M.M. Sousa et al. / International Journal of Biological Macromolecules 80 (2015) 139–148

141

was used to obtain the fiber diameter distributions of each electrospinning trial. 3. Results and discussion 3.1. Solubility of the polymers in the DES

Fig. 1. Schematic representation of the electrospinning setup: (a) syringe pump; (b) tubeless spinneret; (c) coagulating/washing bath; (d) chamber with T control; (e) power supply) adapted from a previous research [18]. Please see Table 1 for detailed information concerning the used electrospinning conditions.

2.2.4. Electrospinning The electrospinning trials were carried out in the same equipment (NaBond Technologies Co., Limited, China) used previously to process agar-in-water [18] but with some modifications (Fig. 1). An ethanol bath with an immersed copper electrode was adapted to the electrospinning set-up (chamber + tubeless spinneret) to be used as collector. The electrode was kept 1.5 cm underneath the ethanol surface in all experiments. The collected samples were left stirring overnight in ethanol and further washed several times with ethanol to ensure that all the DES was removed. This prevented the plasticization of the fibers [8]. Finally, the washed samples were dried under vacuum at 50 ◦ C for a minimum of 24 h. For each spinning solution, several electrospinning conditions were tested including: flow rates (0.5–16 mL/h), voltages (10–30 kV) and distances tip-to-collector (6–12 cm). The best working temperature (50 ◦ C) and distance tip-to-collector (8 cm) were the same as before [18] and were chosen according to preliminary tests, equipment constraints as well as solution properties. The needle size was kept at minimum (outer diameter × length = 0.6 mm × 30 mm and inner diameter = 0.317 mm) to reduce as much as possible the surface area for Taylor cone formation [19].

2.2.5. Scanning electron microscopy (SEM) The morphology and size of the fibers were investigated in a Quanta 200 FEG scanning electron microscope (SEM, FEI, Hillsboro, OR) following a previous procedure [18]. The samples were coated with a thin film of gold after being mounted with adhesive on specimen stubs, and the edge painted with colloidal silver adhesive. The images were acquired in the high-vacuum/secondary electron imaging mode SEM using accelerating voltages close to 10 kV and working distances around 12.5 mm. The average fiber diameters were determined using the image analysis software XT Document (FEI Corp, Hillsboro, OR). Mean values of fiber diameters and respective standard deviations were obtained after measuring a significant number of fiber diameters in the SEM images shown in Figs. 5–9. The Statistica 8.0 software (StatSoft, Tulsa, OK, USA)

Agar and PVA are gelling polymers that swell when dispersed in a compatible solvent. This swelling is caused by polymer–solvent interactions such as hydrogen bonds. At the same time, the many hydroxyl groups present in the polymer chains establish strong intra- and intermolecular hydrogen bonds in solution [29]. These polymer–polymer interactions increase the tendency of both polymers to aggregate and prevent dissolution to occur. To attain optimal dissolution agar and PVA need to be completely dispersed in the solvent and heated [12,18]. Heating causes the disruption of the polymer-polymer hydrogen bonds and consequently increases the polymers’ solubility. Agar and PVA were quickly dissolved in the DES (typically, after ∼20–30 min). PVA particles were easily dispersed in the solvent and did not clump together upon heating. After solubilisation, no undissolved PVA particles (appearing as transparent gels) were visible in the final solution suggesting the complete dissolution of the polymer. Agar was also easily dispersed and dissolved. This fast solubility could be easily explained by the great capacity of the DES to establish hydrogen bonds with the hydroxyl groups present throughout the backbones of agar and PVA (i.e. strong polymer–solvent interactions) [22]. 3.2. Rheological properties of the spinning solutions The first logical step was to analyse the rheological properties of the solvent. At 50 ◦ C, the viscoelastic properties of the DES showed predominance of the viscous character (G > G ), with G describing a steep increase with ω (not shown). The higher viscosity of the DES (around 0.1 Pa s at 50 ◦ C; not shown) when compared to other common solvents (e.g. water), could be attributed to the large ion sizes and very small voids within the DES system as well as attractive forces between its components [22]. A clear Newtonian behavior was also identified from the flow curve for the range of tested shear rates, 1–200 s−1 (not shown). Other relevant properties of the DES have been extensively discussed in previous reference papers [21,22]. In Fig. 2, are reported the cooling ramps from 90 to 50 ◦ C of the pure agar (1 wt%) and pure PVA (5 wt%) as well as agar/PVA blends prepared with the DES. Pure agar-in-DES experienced significant elasticity gain during the cooling process, indicated by a faster increment of G in relation to G (i.e. sharp decrease of the tan ı; not shown). In aqueous media, both moduli could be considered constant at this stage [18]. Assuming similar gelation mechanisms for agar-in-DES and agar-in-water [13,30], this could match the initial formation of double helices for the former and random coil state for the latter. Taking as valid the commonly used rule G = G (or tan ı = 1) defined for gelation, one can see the aggregation process leading to gel formation happening much sooner (i.e. at higher temperature) when agar is dissolved in DES (∼80 ◦ C against ∼30 ◦ C in water [12]). Regarding the agar/PVA mixtures, no transition in the viscoelasticity was observed (i.e. G > G over the entire temperature window) for the 30/70 and 20/80 ratios; oppositely, for the 50/50 ratio, G > G below ∼67 ◦ C (Fig. 2B). This concentration threshold for gelation was not observed when cooling agar/PVA aqueous blends (i.e. G > G in all cases) [18]. The higher number of PVA molecules (non-gelling polymer) in aqueous media (10 wt% PVA starting solution) hindering the formation of agar 3D networks might explain this result. Early studies from Kumar and co-workers described

142

A.M.M. Sousa et al. / International Journal of Biological Macromolecules 80 (2015) 139–148

Fig. 2. Cooling ramp from 90 to 50 ◦ C, at 6.28 rad/s, of agar/PVA blends prepared in DES at different mass ratios (concentrations of starting solutions: 1 wt% agar and 5 wt% PVA).

Fig. 3. Mechanical spectra at 50 ◦ C of agar/PVA blends prepared in DES at different mass ratios (concentrations of starting solutions: 1 wt% agar and 5 wt% PVA).

remarkably lower gelation points (∼17–40 ◦ C) for native and regenerated agarose at 2–5 wt% concentrations, in various room temperature ionic liquids (RTILs) [31,32]. The authors concluded that the presence of large ions would interact with the agarose molecules hindering the double helical conformation. Contrary to conventional ILs, DESs are not entirely formed by ionic species [22]. The equilibration curves at 50 ◦ C confirmed G > G for 30/70, 20/80 and 0/100 agar/PVA, whereas blends with higher agar contents showed G > G (not shown). At the end of this equilibration step, the viscoelasticity of the systems was not undergoing significant changes and so, the mechanical spectra were acquired (Fig. 3). At 50 ◦ C, pure agar-in-DES described typical behavior of ‘true gel’, i.e. G , at least, one-order of magnitude higher than G and both moduli showed little frequency dependence [33]. Extremely low values of viscoelasticity were recorded for agar-in-water for the same conditions [18]. Plotting the mechanical spectra in terms of |* (ω)| logarithmic data (not shown) also proved significant elasticity for pure agar-in-DES, as confirmed by the high values of dynamic consistency index, K, and power law factor, n, respectively, ∼564 Pa s and −0.93 (Eq. (1)). Pure PVA-in-DES behave differently, with G > G in all range of considered ω (Fig. 3A). Agar/PVA blends exhibited intermediate behaviors between those described by the pure polymers. The addition of PVA to the gelled systems had two visible effects, decrease of gel rigidity (G decrease) and decrease of gel elasticity (tan ı increase; not shown). For 30/70 and 20/80 agar/PVA, the behavior of the co-blending polymer becomes clearly predominant with G > G over the entire ω range (Fig. 3B). Accordingly, G is attenuated; for instance,

taking as reference the elastic modulus measured at 6.28 rad/s (G0 ) the following trend could be observed for agar/PVA blends: 100/0 > 50/50 > 30/70 > 20/80 > 0/100 respectively, ∼620, ∼57, ∼6, ∼1.7 and ∼0.86 Pa. Plotting the flow curves of the spinning solutions with predominant viscous component at 50 ◦ C, showed explicitly an increase in apparent viscosity (app ) with PVA addition (Fig. 4). This indicates

Fig. 4. Flow curves at 50 ◦ C of agar/PVA blends prepared in DES at different mass ratios (concentrations of starting solutions: 1 wt% agar and 5 wt% PVA).

A.M.M. Sousa et al. / International Journal of Biological Macromolecules 80 (2015) 139–148

143

Table 1 Electrospinning conditions used to collect the fibers. All experiments were performed at 50 ◦ C, using 3 mL of sample volume and a distance tip-to-collector of 8 cm. Spinning solutions were prepared from a 1 wt% agar and 5 wt% PVA starting solutions. Conditions id

Figure id

Agar/PVA ratio

Flow rate (mL/h)

V (kV)

Observations during trials

EC1 EC2 EC3 EC4 EC5 EC6 EC7

Fig. 5B Fig. 6A/B Fig. 6C/D Fig. 6E/F Fig. 7A/B n.a. Fig. 7C/D; Fig. 8A/B

100/0 50/50 50/50 50/50 30/70 30/70 30/70

5 14 6 6 14 11 6

10a 18 14 18 18 11 20

EC8

Fig. 7E/F

30/70

6

26

EC9

Fig. 8C/D

30/70

6

28

EC10

Fig. 9A/B

20/80

11

11

Continuous and oriented jet Continuous and oriented jet Continuous and oriented jet Continuous and oriented jet Discontinuous and cahotic jet Didn’t form collectable material Continuous, uniform and cahotic jet; two distinct materials collected from the coagulating bathb Continuous, thin and stretched jet; one material collected from the coagulating bathb Discontinuous, short and oriented jet; one electrospun material collected by filtration Continuous and oriented jet; a collectable thin fiber web was formed; very resistant

n.a., not applied. a Increased up to 20 kV during the course of the experiment. b See Section 3 for details.

appreciable structural modifications undergone by the solutions when the co-blending polymer was added. Values of app at shear rates believed to be close to those felt in the needle tip, ∼10 s−1 [34], were 16–20 times higher for agar-in-DES than agar-in-water [18]: ∼6, ∼5, ∼3.8 Pa s for, respectively, 0/100, 20/80 and 30/70 agar/PVA. Differences in the solution and solvent properties could explain these results. Pure PVA solution was shear-thinning with a Newtonian plateau in the lower shear rate range while, for agar/PVA blends, this plateau is ill-defined. Similar weak shear thinning behavior was found when using water as solvent [18]. For the mixed solutions in DES, an increase in viscosity is observed for lower shear rates, the upturn in the flow curve being more pronounced for 30/70 agar/PVA (mixture with higher agar content in Fig. 4), which may be indicative of some incompatibility between both polymers. Roˇsic et al. reported similar behaviors for alginate/PEO and chitosan/PEO aqueous blends [35]. 3.3. Electrospinning Table 1 lists the operational parameters as well as important observations concerning the most relevant electrospinning trials. After several preliminary trials, the distance between the needle tip and the collector was fixed at 8 cm and was the same as used

previously for the electrospinning studies with the aqueous agar solutions [18]. Longer distances were tested, yet the polymer jets described highly unstable paths which, considering the used voltages and electrospinning set-up, could pose safety issues. Other parameters (working T, concentration) were chosen according to the intrinsic properties of the polymers. Due to the used setup (chamber + tubeless spinneret; Fig. 1) as well as the properties of the spinning solutions, the flow rates that could effectively pull out the spinning solution from the needle tip towards the counter electrode were well above the values commonly reported for other natural polymers, respectively, 5–14 mL/h (Table 1) vs ∼0.1–3 mL/h [7,8,36]. As first attempt, we tried to electrospin the 1 wt% agar-inDES solution (Fig. 5A). Far beyond the sol–gel transition (i.e. beginning of the run), the solution possessed good spinnability forming branched nanofibers (indicated by an arrow in Fig. 5B). Viswanathan et al. [7] found similar branched morphologies when electrospinning a 10 wt% cellulose solution in the RTIL 1butyl-3-methylimidazolium chloride. These authors explained the branching of the fibrous material with the high viscosity of the solvent [7]. The polymer jet of the 1 wt% agar-in-DES was continuous and oriented towards the collector (not shown), in clear contrast with the solution droplets observed when using water as solvent

Fig. 5. Photograph of the 1 wt% agar-in-DES solution (A). Representative SEM micrograph at 500× magnification of the pure agar fibers (B) obtained when electrospinning the solution showed in (A). The fiber diameter distribution of the fibers shown in the micrograph is given in (C) (Note: the large fiber agglomerates were not considered for the statistical analysis since they were consequence of process limitations). The arrow indicates an example of a nanofiber.

144

A.M.M. Sousa et al. / International Journal of Biological Macromolecules 80 (2015) 139–148

Fig. 6. Representative SEM micrographs of the 50/50 agar/PVA fibers obtained at 500× (A, C and E) and 2500× (B, D and F) magnifications. The respective fiber diameter distributions are shown as insets. The electrospinning conditions (EC) used to obtain each sample are listed in Table 1 (i.e. A/B—EC2; C/D—EC3; E/F—EC4).

medium [18]. Good conditions for formation of agar nanofibers were a starting voltage of 10 kV and a flow rate of 5 mL/h (EC1 conditions in Table 1). Longer spinning times however, led to an increase in the solution viscosity and gelation (Fig. 2A) [12,13,26], and the electrospinning conditions had to be readjusted. Consequently, the electrospun materials increased in size (up to the sub-micron range as seen in Fig. 5B) ending up as coarse aggregates of fibers when the changes in viscosity and gelation became even more significant. The use of a solvent bath may also increase the tendency for fiber agglomeration during electrospinning due to the slower solvent removal from the fibers when compared to typical electrospinning set-ups dealing with volatile solvents [37]. All these aspects contributed to a broad distribution of fiber diameters (Fig. 5C). At some point, the needle would block and the trial had to be stopped. The spinnability of 1 wt% agar-in-DES was clearly limited by the used equipment and a chamber with a wider range of working temperatures would have allowed a more continuous collection of

agar nanofibers. In our case, the spinning solution was placed inside the chamber at temperatures well above 50 ◦ C (maximum working temperature of the electrospinning set-up) but during the trial, the solution would cool down approaching the temperature set in the equipment (50 ◦ C). Despite this process limitation, it is clear the advantages of electrospinning over other methods to produce fibers with higher surface-to-volume ratio [38]. Next, PVA (5 wt%) was added as a co-blending polymer to further improve agar-in-DES spinnability. 50/50 agar/PVA solution was very easy to process by electrospinning and consistently yielded collectable fibers (Fig. 6). Continuous and well oriented polymer jets were formed at various sets of electrospinning conditions (not shown) suggesting improved viscoelasticity for this blend (Fig. 3A). This good processability was not seen previously when using water as solvent. Here, the aqueous blends with higher contents of agar were hard to process and yielded discontinuous fibrous mats [18]. The polymer jets were also more homogeneous when using DES

A.M.M. Sousa et al. / International Journal of Biological Macromolecules 80 (2015) 139–148

145

Fig. 7. Representative SEM micrographs of 30/70 agar/PVA fibers at 250× (A), 500× (E), 2500× (B and C), and 10,000× (D and F) magnifications. The respective fiber diameter distributions are shown as insets (Note: Large fibers shown in C were not measured since they were too heterogeneous). The electrospinning conditions (EC) used to obtain each sample are listed in Table 1 (i.e. A/B—EC5; C/D—EC7; E/F—EC8).

as solvent medium (not shown) which could indicate more interchain bonds established between agar and PVA in the eutectics. Most of the fibers formed by the 50/50 agar/PVA solution were in the sub-micron range (

Improving agar electrospinnability with choline-based deep eutectic solvents.

Very recently our group has produced novel agar-based fibers by an electrospinning technique using water as solvent and polyvinyl alcohol (PVA) as co-...
3MB Sizes 0 Downloads 20 Views