Rheological Properties and Electrospinnability of High-Amylose

Jul 20, 2015 - (46) who demonstrated that water content strongly influenced the rheological behavior and processability of starch, and was critical in...
0 downloads 11 Views 2MB Size
Article pubs.acs.org/Biomac

Rheological Properties and Electrospinnability of High-Amylose Starch in Formic Acid Anica Lancuški,*,† Gleb Vasilyev,† Jean-Luc Putaux,‡,§ and Eyal Zussman† †

NanoEngineering Group, Faculty of Mechanical Engineering, Technion Israel Institute of Technology, 32000 Haifa, Israel Université Grenoble Alpes, Centre de Recherches sur les Macromolécules Végétales (CERMAV), F-38000 Grenoble, France § CNRS, CERMAV, F-38000 Grenoble, France ‡

S Supporting Information *

ABSTRACT: Starch derivatives, such as starch-esters, are commonly used as alternatives to pure starch due to their enhanced mechanical properties. However, simple and efficient processing routes are still being sought out. In the present article, we report on a straightforward method for electrospinning high-amylose starch-formate nanofibers from 17 wt % aqueous formic acid (FA) dispersions. The diameter of the electrospun starch-formate fibers ranged from 80 to 300 nm. The electrospinnability window between starch gelatinization and phase separation was determined using optical microscopy and rheological studies. This window was shown to strongly depend on the water content in the FA dispersions. While pure FA rapidly gelatinized starch, yielding solutions suitable for electrospinning within a few hours at room temperature, the presence of water (80 and 90 vol % FA) significantly delayed gelatinization and dissolution, which deteriorated fiber quality. A complete destabilization of the electrospinning process was observed in 70 vol % FA dispersions. Optical micrographs showed that FA induced a disruption of starch granule with a loss of crystallinity confirmed by X-ray diffraction. As a result, starch fiber mats exhibited a higher elongation at break when compared to brittle starch films.



INTRODUCTION There is currently a progressive trend to replace synthetic polymers from limited petroleum-based resources with sustainable natural macromolecules and their derivatives.1−5 Starch is a natural, abundant polysaccharide mainly present in plants, where it serves as energy storage. It occurs as granules whose size depends on the botanical origin (typically 1−100 μm) that are insoluble in water at room temperature. Starch is a mixture of amylose, a mostly linear polymer of α(1 → 4)-linked glucosyl units (typically 20−30 wt % of standard starch) and amylopectin constituted of linear segments connected by 5−6% α(1 → 6) branch points (70−80 wt % of standard starch).6 The starch granules are semicrystalline, and the crystallinity index depends on the water content. Due to its clustered organization of short branches, associated into double helices, amylopectin is believed to be mostly responsible for the overall crystallinity. Starch granules can be extracted from various sources such as corn, wheat, potato, tapioca, or rice, each containing different ratios of amylose and amylopectin. A special type of highcontent (up to 70%) amylose corn starch, referred to as Hylon starch, adds to versatility of physicochemical properties and applications of the starch. The advantage of using high-amylose starch lies in its superior strength and toughness in the preparation of starch-based materials and in the possibility to produce various modified starches by reactive extrusion.7,8 As a “generally recognized as safe” (GRAS) material, approved by the Food and Drug Administration (FDA), starch has been extensively used in medical, pharmaceutical, and food industries © XXXX American Chemical Society

as an inexpensive support for drug delivery and in food packaging.9 However, pure starch materials, typically starch films, are rather brittle, water-sensitive, and difficult to be processed,10,11 which limits their applications. In order to overcome the brittleness of the starch-based materials, additives, such as salts, to eliminate the water-evaporation process, or a polyacrylamide polymer or its derivatives, to reduce the shear viscosity and increase the extensional viscosity, have been used in extrusion and melt-spinning procedures.12,13 Furthermore, orientation of starch in extruded and stretched fibers results in enhancement of tensile strength and elongation at break.14 Kong and Ziegler15−17 studied the processing of pure starch fibers by electrospinning from a dimethyl sulfoxide (DMSO)-rich solvent medium,18−21 and established the relationship between the electrospinnability and rheological properties of starch (such as entanglement concentration and shear viscosity), processing parameters (such as spinning distance, flow rate, and voltage) and the amylose/amylopectin ratio. Due to fiber brittleness, and in order to improve the crystallinity and water-stability of the fibers, different postprocessing treatments, like annealing and cross-linking, were applied,22 and the formation of starch-guest inclusion complexes was investigated as well.23,24 Xu et al.25 reported on electrospinning starch-acetate nanofibers from an aqueous Received: June 18, 2015 Revised: July 17, 2015

A

DOI: 10.1021/acs.biomac.5b00817 Biomacromolecules XXXX, XXX, XXX−XXX

Article

Biomacromolecules

carbon tape, were coated with gold using an Emitech sputter coater for 15 s, and then observed in secondary electron mode, with a Phenom microscope operating at a partial pressure 0.01 mbar, an electron acceleration voltage of 5 kV. Optical Microscopy. Dispersions of HS in FA (17 wt %) were observed with an Olympus BX51 optical microscope equipped with an Olympus DP12 camera, and a Zeiss Axiophot II microscope equipped with polarized light illumination and a SIS ColorView 12 camera. Rheological Measurements. A Discovery DHR-2 rotational rheometer (TA Instruments, USA) was used to characterize the rheological properties of solutions under steady-state shear flow. Parallel-plate geometry, with a diameter of 40 mm and a gap of 0.5 mm, was applied. Oscillatory shear deformation tests were carried out in the region of linear viscoelastic response of the materials. All rheological measurements were performed at room temperature (25 °C). X-ray Diffraction (XRD). For XRD analysis, starch-formate fibers were specifically spun on a synchronized rotating drum in order to obtain films with a high fibrillar orientation. Thin strips of the films were cut and placed into 1 mm-wide glass capillaries. Some specimens were placed in a 95% relative humidity environment and left for 1 week. All capillaries were flame-sealed and X-rayed in a vacuum chamber using a Philips PW3830 generator operating at 30 kV and 20 mA (Ni-filtered CuKα radiation, λ = 0.1542 nm), in transmission mode. The film strips were oriented either parallel or perpendicular to the X-ray beam. Two-dimensional diffraction patterns were recorded on Fujifilm imaging plates. The plates were then read off-line, using a Fujifilm BAS 1800-II bioimaging analyzer. Mechanical Analyses. The mechanical properties of the electrospun mats were investigated by tensile testing at room temperature, using a Dynamic Mechanical Analyzer (DMA, Q800-TA Instruments, USA). Samples were approximately 20 mm-long, 5 mm-wide, and 0.15−0.25 mm-thick. Stress−strain curves were recorded at a stretching rate of 1%/min. Stress was calculated according to the effective area of the sample, σ = F/(A·ρmat/ρbulk), where F is the measured force, A is the measured cross-section, the apparent density is ρmat/ρbulk, where ρbulk is the starch tapped powder density, and ρmat is the fiber mat density.

formic acid (FA) solution. The tensile strength of the nanofibers was found to depend on the starch−acetate concentration, annealing time, and degree of substitution (DS). It is well documented that in FA, starch undergoes a rapid esterification, called o-formylation.26−29 The action of FA on starch at room temperature is regioselective, generating monoformate esters at the C6 position of the glucose units of starch,27 and reaches equilibrium after ∼8 h in 90% formic acid solution.29 The reversibility of the starch esterification is exploited in the preparation of oriented starch films.8,30 The oformylation of starch is also known to promote mixing with other biodegradable polymers.31 Moreover, FA results from the γ irradiation of starch and represents the main part of radioinduced free acidity.32 The hydrothermal gelatinization of starch has been extensively studied.33−37 In excess water, and above the gelatinization temperature, the starch granules swell. Amylose segregates from amylopectin, and the crystallites melt by disentanglement of double helices.38 These complex changes in the macromolecular organization strongly depend on the temperature, water content, and amylose/amylopectin ratio as well as on the difference in solubility of these polymers. However, FA and certain salt solutions (NaCl, CaCl2, KSCN, etc.)39 can promote starch gelatinization even at room temperature, in a process known as chemical gelatinization. 40−42 Divers et al. 43,44 investigated the chemical modification and gelatinization process of wheat starch in pure and dilute aqueous FA and noted that lower FA concentrations resulted in increased gelatinization temperatures, while the DS of the starch decreased. However, the origin of the complex rheological behavior of starch in FA remained unclear. In this work, we studied chemical gelatinization, namely, the process of intermolecular bond breakdown in native starch granules, in pure and dilute aqueous FA, as well as the ability of the dissolved and esterified starch in FA to form fibers via electrospinning. First, electrospinnability was investigated as a function of FA/water ratio. Then, the morphological evolution of native starch granules in pure and aqueous FA and the rheological properties of the resulting starch-formate solutions were investigated and discussed in correlation with electrospinnability. Finally, the structural characteristics and mechanical properties of mats of electrospun fibers were studied.





RESULTS AND DISCUSSION

Electrospinning of Hylon VII Starch-Formate from Formic Acid Dispersions. HS dissolved in various concentrations of FA was partially esterified into starch-formate27,29 (see Supporting Information, Figure S1) and successfully electrospun. Our preliminary attempts at electrospinning showed that the 17 wt % HS dispersion was optimal in terms of process stability and fiber uniformity (Supporting Information, Figures S2 to S4). In comparison, Xu et al.25 reported that 20 wt % solutions of starch−acetate in 90 vol % FA resulted in uniform fiber formation, while, Kong and Ziegler showed that uniform fibers could be obtained for 10−20 wt % starch dispersions in aqueous DMSO.16 Fibers were only formed from the HS-FA100, HS-FA90, and HS-FA80 dispersions (Figure 1). HS-FA100 and HS-FA90 dispersions yielded uniform fibers (Figure 1a,b), while HSFA80 resulted in slightly beaded fibers (Figure 1c). When working with solutions with lower FA content (HS-FA70 and HS-FA60), the processing of starch-formate under a high electric field resembled a combination of electrospinning and electrospraying (data not shown), yielding short nanofibers frequently accompanied by beads-on-string.45 It was noted that the diameter of the fibers decreased with increasing water content in FA. For HS-FA100 fibers, the average diameter was 304 ± 53 nm, while for the HS-FA90 fibers, diameter decreased to 156 ± 33 nm. A further decrease in FA content resulted in fibers with a 84 ± 21 nm diameter,

MATERIALS AND METHODS

Materials. High-amylose Hylon VII maize starch (HS) was obtained from National Starch & Chemical, U.K. The amylose content was 70%, as determined by the provider. Formic acid (FA) (98% purity) was purchased from Sigma. All materials were used as received, without further purification. Electrospinning. HS (17 wt %) was dissolved in different dilutions of FA in water with continuous stirring and electrospun at room conditions of temperature and humidity (ranging from 40 to 60%). The 17 wt % starch dispersions in pure, 90, 80, 70 and 60% FA will be referred to as HS-FA100, HS-FA90, HS-FA80, HS-FA70, and HSFA60, respectively, in the following. Electrospinning was performed after 2, 6, 24, and 72 h for HS-FA100, HS-FA90, HS-FA80, and HSFA70 dispersions, respectively, after a complete dissolution took place. The polymer dispersions were poured in a plastic syringe with a G23 blunt needle, and electrospun at an 11 cm distance from the collector, under a flow rate of 0.3 mL/h and a high voltage of 17 kV. A rotating disc with diameter of 8 cm, and rotating at the speed of 3400 rpm was used as a collector to align the fibers. Scanning Electron Microscopy. Samples of electrospun fibers, mounted on SEM stub using a double-stick electrically conductive B

DOI: 10.1021/acs.biomac.5b00817 Biomacromolecules XXXX, XXX, XXX−XXX

Article

Biomacromolecules

temperature, in a similar fashion to thermally induced gelatinization of starch in water.41,42 However, very few studies have considered chemical gelatinization of starch in FA.43,44 In the following, as it is known that o-formylation of starch takes place shortly after dissolution in FA, the starch in FA will be referred to as starch-formate. The physicochemical and morphological changes of Hylon VII starch granules in aqueous FA were followed by optical microscopy in bright-field mode and under polarized light. Figure 2 compares the morphology of starch granules initially

Figure 1. SEM images of Hylon VII starch-formate fibers electrospun from solutions containing (a) pure (HS-FA100), (b) 90 vol % (HSFA90), and (c) 80 vol % (HS-FA80) formic acid.

bearing a micron-sized bead appearance (HS-FA80). These observations suggest a strong influence of water on fiber formation. With increased water content in the solution, the fiber quality dramatically deteriorated. These results corroborate with those reported by McGrane et al.46 who demonstrated that water content strongly influenced the rheological behavior and processability of starch, and was critical in starch gelatinization (granule swelling) and gelation (formation of a network between free amylose and amylopectin). The authors also showed that, upon progressive replacement of DMSO with water, intramolecular bonds were replaced with intermolecular hydrogen bonds, and viscoelastic liquids were transformed into strong gels.46 Similarly, in the case of wheat starch in FA, Divers et al.43,44 demonstrated that the rheological behavior of starch-formate strongly depended on the water/FA ratio. While wheat starch granules underwent complete disruption without hydrolysis when dissolved in pure FA, the starch granules exhibited only a slight swelling in 40% aqueous FA solutions. In the following sections, we present further results on the behavior of Hylon VII starch granules and starch-formate in aqueous FA solutions, obtained from optical microscopy observations and rheological studies. Chemical Gelatinization, Gelation, and Phase Separation of Starch-Formate in FA. During gelatinization, starch granules swell, and lose their optical birefringence and crystallinity. These phenomena are associated with amylose leaching, followed by dissolution, gelation, and finally phase separation of amylose and amylopectin.38,47−50 It is known that different salt and alkaline media can gelatinize starch at room

Figure 2. Optical micrographs of Hylon VII starch granules in water (a−c) and dispersed for 20 min in 70 vol % FA (d−f), and 10 min in 100 vol % FA (g−i) at room temperature. Images a, d, and g were recorded in bright field mode while images b, c, e, f, h, and i are polarized light micrographs. For images c, f, and i, a λ retardation plate was used to generate polarization colors. The white arrowhead in e points at a strongly birefringent granule unaffected by FA after 20 min.

dispersed in water with that after a short incubation in 70 and 100 vol % FA, at room temperature. Native Hylon VII granules are polygonal distorted spherulites (Figure 2a) that showed a clear extinction Maltese cross when observed between crossed nicols (Figure 2b). The use of a λ retardation plate allowed the generation of polarization colors whose distribution in the granules agrees with the well-known spherulitic radial organization of the macromolecules with respect to the granule hilum (Figure 2c).6 The chemical gelatinization kinetics was found to be influenced by the FA concentration. For the dispersion in pure FA, the chemical gelatinization of HS was C

DOI: 10.1021/acs.biomac.5b00817 Biomacromolecules XXXX, XXX, XXX−XXX

Article

Biomacromolecules

Figure 3. Optical micrographs of Hylon VII starch granules dispersed in 70 vol % FA: (a−c) after 30 s; (d−f) after 15 min; (g−i) after 30 min. Images a, d, and g were recorded in bright field mode, while images b, c, e, f, h, and i are polarized light micrographs. For images c, f, and i, a λ retardation plate was used to generate polarization colors. The white arrowheads in b point at granules whose birefringence does not seem to be affected by FA during the 30 min treatment. Their breaking into smaller fragments is likely due to an increase of pressure of the coverslip during the partial evaporation of the liquid.

not be successful. In addition, long dissolution times that resulted in starch phase separation and aggregation may be considered nonspinnable, due to the presence of micron-sized particles.51 Similarly, for electrospinning of PEO-(PEGHMWH) hydrogels, Schultz et al.52 evidenced the importance of establishing a time window of electrospinnability of the hydrogel, during its transition from a sol to a gel. To better assess the spinnability of the starch/FA/water systems, we investigated their morphological evolution and viscoelastic behavior with time. Typical changes in storage (G′) and loss (G″) moduli of the HS-FA70 dispersion as a function of frequency and time are shown in Figure 5. During the first stage of starch swelling, the granule size increases, and the resulting so-called “ghosts” form a gel-like structure (Figure 4b). At this stage, dynamic moduli as a function of frequency reveal a solid-like behavior with G′ > G″ (Figure 5, 10 h). Both moduli were almost independent of the applied frequency, with the G′ plateau value reflecting the strength of the structure formed. As granule swelling progressed, amylose leached from the granule. The gel strength weakened, and a gradual transition from solid-like to liquid-like behavior occurred with G″ > G′ (Figure 5, 24 h). When the HS granules were completely disrupted, starch-formate dissolution took place (Figure 4c) and the solution showed typical liquidlike behavior (Figure 5, 3 days). The slope of the G″(ω) curve in the region of low frequencies is close to the theoretical one

very fast, and the optical birefringence was lost within a few seconds (Figure 2g−i). However, in 70 vol % FA, for the same observation time, the chemical gelatinization was significantly slower (Figure 2d−f). To slow down the gelatinization kinetics and facilitate the time-resolved observation, a dispersion of HS in 70 vol % FA was used (Figure 3). HS granules exhibited a rather low swelling capacity, which began at the center of the granule.48 The birefringence of HS granules was lost within 10 min, although the swelling was not extensive. A small fraction of highly birefringent granules seemed to be resistant to the chemical gelatinization within the observation period. Their subsequent fragmentation was likely due to the increase of the pressure exerted by the coverslip during liquid evaporation. When monitored by optical microscopy, the time-resolved effect of FA on native HS granules (Figure 4a) progressed in three stages: (i) starch granule swelling and simultaneous oformylation (Figure 4b), (ii) starch-formate dissolution (Figure 4c), and (iii) phase separation and aggregation (Figure 4d). The starch concentration in FA selected for this analysis (namely, 17 wt %) was that used for the electrospinning purpose (see Supporting Information, Figures S2 to S4). The solution evolved from gel-like to liquid-like and finally to a twophase system. Gel-like solutions are known to be not spinnable, due to their high viscosity. It could therefore be suspected that attempts to electrospin HS at an early stage of swelling would D

DOI: 10.1021/acs.biomac.5b00817 Biomacromolecules XXXX, XXX, XXX−XXX

Article

Biomacromolecules

point, HS-FA100, HS-FA90, and HS-FA80 dispersions were successfully electrospun into nanofibers. However, the HSFA70 dispersion was not electrospinnable at this stage, possibly arising from the fact that 70% FA was not able to completely disentangle amylose from amylopectin and provide enough leached amylose for the strong entanglement network formation required for continuous electrospinning. Xie et al.53 considered that, for the same concentration of starch in aqueous dispersions and under all the conditions, only linear amylose was contributing to the entanglements formation, while amylopectin formed gel-balls and globular structures that did not contribute in the (effective) entanglement. The decrease in viscosity and moduli of aqueous starch dispersions, resulting from the decrease in entanglement network density due to the decrease in amylose content, may explain in a similar fashion the loss of electrospinnability of HS-FA70 dispersions. A temperature decrease after thermal gelatinization of starch in excess water is known to promote reorganization of amylose and amylopectin in the solution and further recrystallization.54 Similarly, during the chemical gelatinization in FA, starch undergoes macromolecular reorganization with time at room temperature. Longer storage times induced phase separation and further aggregation of the starch (Figure 4d). The aggregate size increased with time as a result of polymer precipitation and consolidation of small aggregates. Finally, eyevisible particles could be observed in the solution and the dispersion demonstrated a pseudosolid-like behavior (Figure 5, 8 days). In this case, G″ was higher than G′, but the elastic modulus plateau appeared in the region of low frequencies, reflecting strong structuration in the system. This behavior is typical for filled polymer melts and solutions.55 Decreased G′ values at high frequencies reflect weakening of the formed entanglement network, which can be attributed to a concentration drop of the dissolved polymer fraction. Figure 6 shows the evolution of shear viscosity (η) and elastic modulus (G′) with time for the 17 wt % HS dispersions. The permanent decrease of η and G′ over time was observed for all tested dispersions. The time-dependent behavior of η and G′ could be divided into two distinct kinetic patterns: (i) a rapid decrease in viscosity, resulting from o-formylation, granule swelling and disruption, and simultaneous dissolution of starchformate in FA29 (Figure 4b,c), and (ii) small changes in viscosity, most likely due to macromolecular reorganization and phase separation,56 as previously observed (Figure 4d). By increasing the water content, the onset time where the kinetics of the system changes, increases dramatically. In pure formic acid, o-formylation and dissolution of starch-formate seemed to occur within the first 2 h, justifying the absence of the sharp drop in viscosity. In contrast, increased water content in HS− FA dispersions considerably decelerated granule swelling and disruption. In addition, the transition from amylose leaching to complete dissolution occurred after 6, 24, and ∼72 h after preparing the HS-FA90, HS-FA80, and HS-FA70 dispersions, respectively. For the HS-FA60 dispersion, only swollen granules were observed. Optical microscopy observations showed that the complete dissolution did not take place even after 8 days (data not shown). Extended aging periods further promoted phase separation and strong aggregation, leading to increased G′ in the region of low frequency (Figure 5). As rheological studies suggest, starch-formate dispersions in FA are likely to age, and for the same starch concentration (17 wt %), different FA/water compositions show different time windows for successful electrospinning. Very close absolute values for η

Figure 4. Photographs of the vials, optical micrographs (scale bars: 50 μm), and corresponding schemes of typical stages in the swellingdissolution−precipitation process of 17 wt % starch in FA: (a) initial state: HS granules were dispersed in water and they are schematized with their Maltese cross birefringence pattern; (b) swelling resulting in the formation of ghosts; (c) dissolution/chain entanglement. The apparent turbidity in b and c is very similar, but both dispersions are very different in terms of rheology: in b, it is gel-like, while in c, it is liquid-like; (d) phase separation and aggregation observed after a 1month storage.

Figure 5. Variation of storage (filled symbols) and loss modulus (open symbols) with increasing frequency for the HS-FA70 dispersion, at different time intervals and at 25 °C.

(G″ ∼ ω1). However, the slope of the G′(ω) curve is lower than the theoretical one (G′ ∼ ω2) and close to that of the G″(ω) curve, meaning that a very weak structuration takes place in the solution. Divers et al.44 alluded to the fact that, in the first stages of dissolution, FA does not destroy the helical structure of starch. Thus, the weak structuration observed at this stage may be ascribed to helical entities present in the solution that led to the change of the curve’s slope. At this E

DOI: 10.1021/acs.biomac.5b00817 Biomacromolecules XXXX, XXX, XXX−XXX

Article

Biomacromolecules

Figure 7. Flow curves of (a) viscosity as a function of shear rate and (b) storage modulus as a function of frequency for the solutions (1) HS-FA100 after 4 h, (2) HS-FA90 after 10 h, (3) HS-FA80 after 24 h, and (4) HS-FA70 after 72 h of dissolution at 25 °C.

Figure 6. Time-dependence curves of (a) shear viscosity η and (b) storage modulus G′ for HS-FA100 (1), HS-FA90 (2), HS-FA80 (3), HS-FA70 (4) and HS-FA60 (5) dispersions at γ̇ = ω = 0.4 rad/s and at 25 °C. Hatched rectangles represent the time range of viscosities and storage moduli of electrospinnable HS-FA dispersions.

dispersion were significantly lower than the rest of the solutions throughout the entire range of applied frequencies. Elasticity of this dispersion dropped more significantly than its related viscosity in comparison to the HS-FA100 dispersion, and approximately 4-fold G′ decrease was registered. Apparently, this significant drop of elasticity reflecting the sparseness of the entanglement network was enough to preclude spinnability of HS-FA70 solution. Structural Analysis of Fibers. The electrospun fibers were characterized by X-ray diffraction (XRD). Figure 8a shows the two-dimensional XRD pattern of the initial hydrated semicrystalline HS granules. It corresponds to that of allomorph B, as expected for high-amylose starch granules and in agreement with previous analyses of Hylon VII starch.59,60 Figures 8b and 8c show the patterns collected from mats of parallel fibers, oriented parallel and perpendicular to the X-ray beam, respectively. In both cases, only one main diffuse ring is observed, which is consistent with amorphous starch-formate fibers. In addition, although the electrospinning process induces a strong elongation flow and high strain rates,57 the homogeneous intensity along the diffusion rings attests that there is no preferential orientation of the macromolecules in the fibers. There was no significant difference after a long hydration period (several days), suggesting that no retrogradation of starch-formate occurred (data not shown). Mechanical Analysis. Stress−strain curves of electrospun mats derived from different FA compositions are shown in Figure 9. The stress was determined based on the apparent density of the fibrous mat. For each type of fibrous mat, the maximum stress (σmax), elongation at break (ε*) and Young’s

and G′ were observed and highlighted in Figure 6. It was observed that electrospinnable HS−FA dispersions lie at the crossover point of two distinct kinetics observed for η and G′, and have the viscosities (η) in the region of 1 Pa·s and elastic modulus (G′) in the region of 0.1 Pa (Figure 6, hatched rectangles). It should be noted that for the HS-FA70 dispersion, these values, at around 72 h, do not fall within the hatched region and, in addition, do not result in fiber formation under a high electric field. Only an unstable process yielding short fibers and beads was observed. Figure 7 presents flow curves and storage moduli versus frequency dependencies for 17 wt % HS dispersions in different FA concentrations. All measurements were taken after granule disruption and starch-formate dissolution. It was evident that the HS-FA70 dispersion had a much lower viscosity and storage modulus than the other dispersions. In order to find the shear viscosity at 1000 rad/s,57 a common shear rate generated during electrospinning, the Sisko model,58 which is useful to describe flow properties of shear-thinning materials at high shear rates, was employed. Fitting the curves results in the viscosity values of 0.79, 0.74, 0.47, and 0.25 Pa·s for HS-FA100, HS-FA90, HS-FA80 and HS-FA70, respectively. The viscosity decreased by about 40% for HS-FA80 dispersion, and only beaded fibers could be obtained. For the unspinnable HS-FA70 dispersion, the viscosity dropped significantly, to approximately 3-fold that of the HS-FA100 dispersion. The same tendency was observed for the storage moduli (Figure 7b). The values of storage modulus of the HS-FA70 F

DOI: 10.1021/acs.biomac.5b00817 Biomacromolecules XXXX, XXX, XXX−XXX

Article

Biomacromolecules

times for the HS-FA80 fiber mats (σmax = 7.01 MPa, ε* = 6.7% and E0.5 = 234 MPa). When comparing to the average diameters of the tested fibers, it can be noted that the fibers with the largest diameter (HS-FA100) had the highest maximum stress, elongation at break, and Young’s modulus. HS-FA90 electrospun fibers showed slightly lower values for σmax, ε* and E0.5. HS-FA80 electrospun fibers showed a gentle decrease in σmax and E0.5 compared to the other fiber mats, but a significant 4-fold decrease in elongation when compared to fibers electrospun from pure FA dispersions. The decrease in mechanical properties with increasing water content contributes to validate the hypothesis that with increasing water content, the fraction of free amylose decreases, further decreasing the entanglement network density and resulting in thinner fibers and deterioration of the overall quality of the fibers. On the one hand, it could be observed that Young’s modulus and tensile strength of electrospun fibers were significantly lower than those of the high-amylose starch films. This might be due to the loss of crystallinity in electrospun fibers. On the other hand, elongation at break was notably higher for electrospun fibers and decreased toward the value for high-amylose starch film as the formic acid concentration in the dispersion decreased.

Figure 8. 2-D X-ray diffraction diagrams of (a) hydrated Hylon VII starch granules, (b,c) a mat of parallel electrospun fibers (HS-FA80), oriented parallel (b) and perpendicular (c) to the beam. The fibers are oriented vertically with respect to the diagrams.

modulus at elongation of ε = 0.5% (E0.5) were extracted (Table 1). Table 1. Mechanical Properties of Native Hylon VII Starch Film and Starch-Formate Electrospun Fibers

HS-FA100 fibers HS-FA90 fibers HS-FA80 fibers Hylon VII cast film61

σmax (MPa)

ε* (%)

E0.5 (MPa)

ρ (g/ cm3)

9.38 ± 0.4 7.01 ± 0.1 5.60 ± 0.5 40 ± 13

26 ± 5.0 21 ± 2.0 6.7 ± 0.3 1.92 ± 1.0

264 ± 37 234 ± 9 218 ± 18 3390 ± 387

0.652 0.633 0.605 0.730

CONCLUSIONS



ASSOCIATED CONTENT

This Article reports on a one-pot method for starch gelatinization and processing into electrospun fibers from formic acid solutions. FA played a crucial role by disrupting the starch granule structure, solubilizing and esterifying starch into starch-formate, and acting as a dispersing medium for electrospinning. Electrospinning of starch-formate from HSFA100, HS-FA90, and HS-FA80 dispersions resulted in the formation of fibers with diameters ranging from 80 to 300 nm. The optimal electrospinning conditions for these solutions were time-dependent, occurring after the complete granule disruption and solubilization but before phase separation and aggregation took place. Rheological measurements and optical microscopy imaging showed that starch gelatinization kinetics in FA was dependent on the water content. Progressive increase of the water fraction in FA delayed the gelatinization and solubilization of starch, decreasing the degree of entanglement in the solution and thereby considerably deteriorated the quality of the electrospun fibers, as manifested by a significant decrease in the elongation at break and tensile strength. However, the resulting fibers, which appeared to be amorphous and without preferential molecular orientation, exhibited higher elongation at break when compared to native starch films, highlighting the potential of electrospun starch fibers as a lowcost and sustainable biomaterial, applicable in the food packaging or pharmaceutical industries.

Figure 9. Stress (σ)−strain (ε) curves of electrospun starch-formate fibers prepared from HS-FA100, HS-FA90, and HS-FA80 dispersions.

sample



Overall, the mechanical properties of the fibers decreased with increased water content during the electrospinning solution. While the HS-FA100 fiber mats featured high maximum stress, elongation at break and Young’s modulus (σmax = 9.38 MPa, ε* = 26% and E0.5 = 264 MPa), these parameters were significantly decreased; σmax and E0.5 for the HS-FA90, while the value of elongation at break was reduced 4

* Supporting Information S

FTIR-ATR spectra of starch-formate from FA and pure starch from DMSO cast films and flow curves of HS-FA100, HSFA90, and HS-FA80 dispersions are shown. The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.biomac.5b00817. G

DOI: 10.1021/acs.biomac.5b00817 Biomacromolecules XXXX, XXX, XXX−XXX

Article

Biomacromolecules



(23) Kong, L.; Ziegler, G. R. Carbohydr. Polym. 2014, 111, 256−263. (24) Kong, L.; Ziegler, G. R. Food Hydrocolloids 2014, 38, 211−219. (25) Xu, W.; Yang, W.; Yang, Y. Biotechnol. Prog. 2009, 25 (6), 1788−1795. (26) Alexander, J. J. Soc. Chem. Ind., London 1936, 55, 206−209. (27) Gottlieb, D.; Caldwell, C. G.; Hixon, R. M. J. Am. Chem. Soc. 1940, 62, 3342−3344. (28) Wolff, I. A.; Olds, D. W.; Hilbert, G. E. J. Am. Chem. Soc. 1951, 73, 346−349. (29) Wolff, I. A.; Olds, D. W.; Hilbert, G. E. J. Am. Chem. Soc. 1957, 79, 3860−3862. (30) Shogren, R. Biomacromolecules 2007, 8, 3641−3645. (31) Pillin, I.; Divers, T.; Feller, J.-F.; Grohens, Y. Macromol. Symp. 2005, 222, 233−238. (32) Dauphin, J. F.; Athanassiadis, H.; Berger, G.; Saint-Lèbe, E. L. Starch - Stärke 1974, 26, 14−17. (33) Tako, M.; Hizukuri, S. J. Carbohydr. Chem. 1999, 18, 573−584. (34) Tako, M.; Hizukuri, S. Carbohydr. Polym. 2002, 48, 397−401. (35) Liu, H.; Lelievre, J.; Ayoung-Chee, W. Carbohydr. Res. 1991, 210, 79−87. (36) Tako, M. J. Appl. Glycosci. 2000, 47, 187−192. (37) Miles, M. J.; Morris, V. J.; Ring, S. G. Carbohydr. Res. 1985, 135, 257−269. (38) Keetels, C. J. A. M.; van Vliet, T.; Walstra, P. Food Hydrocolloids 1996, 10, 343−353. (39) Jane, J.-L. Starch - Stärke 1993, 45, 161−166. (40) Aburto, J.; Hamaili, H.; Mouysset-Baziard, G.; Senocq, F.; Alric, I.; Borredon, E. Starch - Stärke 1999, 51, 302−307. (41) Pan, D. D.; Jane, J. Biomacromolecules 2000, 1, 126−132. (42) Kuakpetoon, D.; Wang, Y.-J. Carbohydr. Res. 2007, 342, 2253− 2263. (43) Divers, T.; Pillin, I.; Feller, J.-F.; Levesque, G.; Grohens, Y. Starch - Stärke 2004, 56, 389−398. (44) Divers, T.; Balnois, E.; Feller, J.-F.; Spevacek, J.; Grohens, Y. Carbohydr. Polym. 2007, 68, 136−145. (45) Greenfeld, I.; Zussman, E. J. Polym. Sci., Part B: Polym. Phys. 2013, 51, 1377−1391. (46) McGrane, S. J.; Mainwaring, D. E.; Cornell, H. J.; Rix, C. J. Starch - Stärke 2004, 56, 122−131. (47) Jenkins, P. J.; Donald, A. M. Int. J. Biol. Macromol. 1995, 17, 315−321. (48) Bear, R. S.; Samsa, E. G. Ind. Eng. Chem. 1943, 35, 721−726. (49) Keetels, C. C. J. A. M.; van Vliet, T.; Walstra, P. Food Hydrocolloids 1996, 10, 355−362. (50) Keetels, C. J. A. M.; van Vliet, T.; Walstra, P. Food Hydrocolloids 1996, 10, 363−368. (51) Daga, V. K.; Helgeson, M. E.; Wagner, N. J. J. Polym. Sci., Part B: Polym. Phys. 2006, 44, 1608−1617. (52) Schultz, K. M.; Campo-Deaño, L.; Baldwin, A. D.; Kiick, K. L.; Clasen, C.; Furst, E. M. Polymer 2013, 54, 363−371. (53) Xie, F.; Yu, L.; Su, B.; Liu, P.; Wang, J.; Liu, H.; Chen, L. J. Cereal Sci. 2009, 49, 371−377. (54) Tako, M.; Tamaki, Y.; Teruya, T.; Takeda, Y. Food Nutr. Sci. 2014, 05, 280−291. (55) Vasilyev, G. B.; Makarova, V. V.; Rebrov, A. V.; Picken, S. J.; Smirnova, N. M.; Malkin, A. Y.; Kulichikhin, V. G. J. Appl. Polym. Sci. 2011, 120, 3642−3653. (56) Kalichevsky, M. T.; Ring, S. G. Carbohydr. Res. 1987, 162, 323− 328. (57) Reneker, D. H.; Yarin, A. L. Polymer 2008, 49, 2387−2425. (58) Sisko, A. W. Ind. Eng. Chem. 1958, 50, 1789−1792. (59) Khachatryan, G.; Krzeminska-Fiedorowicz, L.; Nowak, E.; Fiedorowicz, M. LWT - Food Sci. Technol. 2014, 58, 256−262. (60) Luo, Z.-G.; Shi, Y.-C. J. Agric. Food Chem. 2012, 60, 9468−9475. (61) Koch, K.; Gillgren, T.; Stading, M.; Andersson, R. Int. J. Biol. Macromol. 2010, 46, 13−19.

AUTHOR INFORMATION

Corresponding Author

*Address: NanoEngineering Group, Faculty of Mechanical Engineering, Technion Israel Institute of Technology, 32000 Haifa, Israel. E-mail: [email protected]. Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS We gratefully acknowledge the financial support of the RBNIRussell Berrie Nanotechnology Institute, and the National Research Foundation (R-398-001-065-592) of Singapore within the framework of the Regenerative Medicine Initiative. The authors would also like to thank Asher Shazman, Ory Ramon and Henri Chanzy for helpful discussions. A.L. acknowledges the support of the Technion postdoctoral fellowship.



ABBREVIATIONS DS, degree of substitution; FA, formic acid; HS-FA100, 17 wt % solution of starch in pure formic acid; HS-FA90, 17 wt % solution of starch in 90 vol % formic acid; HS-FA80, 17 wt % solution of starch in 80 vol % formic acid; HS-FA70, 17 wt % solution of starch in 70 vol % formic acid; HS-FA60, 17 wt % solution of starch in 60 vol % formic acid



REFERENCES

(1) Lee, K. Y.; Jeong, L.; Kang, Y. O.; Lee, S. J.; Park, W. H. Adv. Drug Delivery Rev. 2009, 61, 1020−1032. (2) Stijnman, A. C.; Bodnar, I.; Hans Tromp, R. Food Hydrocolloids 2011, 25, 1393−1398. (3) Elsabee, M. Z.; Naguib, H. F.; Morsi, R. E. Mater. Sci. Eng., C 2012, 32, 1711−1726. (4) Xie, F.; Halley, P. J.; Avérous, L. Prog. Polym. Sci. 2012, 37, 595− 623. (5) Nseir, N.; Regev, O.; Kaully, T.; Blumenthal, J.; Levenberg, S.; Zussman, E. Tissue Eng., Part C 2013, 19, 257−264. (6) Pérez, S.; Bertoft, E. Starch - Stärke 2010, 62, 389−420. (7) Van Soest, J. J. G.; Borger, D. B. J. Appl. Polym. Sci. 1997, 64, 631−644. (8) Yu, L.; Christie, G. J. Mater. Sci. 2005, 40, 111−116. (9) Le Corre, D.; Bras, J.; Dufresne, A. Biomacromolecules 2010, 11, 1139−1153. (10) Sajilata, M. G.; Singhal, R. S.; Kulkarni, P. R. Compr. Rev. Food Sci. Food Saf. 2006, 5, 1−17. (11) Bagliotti Meneguin, A.; Stringhetti Ferreira Cury, B.; Evangelista, R. C. Carbohydr. Polym. 2014, 99, 140−149. (12) Eden, J.; Trksak, R. M. Process for spinning starch fibers. U.S. Patent US4853168 A, August 1, 1989. (13) Bailey, V. A.; Mackey, L. N.; Trokhan, P. D. High molecular weight polymer forms effective entanglements or associations with neighboring starch molecules; melt spinning. U.S. Patent US6709526 B1, March 23, 2004. (14) Wolff, I. Ind. Eng. Chem. 1958, 50, 1552−1552. (15) Kong, L.; Ziegler, G. R. Recent Pat. Food Nutr. Agric. 2012, 4, 210−219. (16) Kong, L.; Ziegler, G. R. Biomacromolecules 2012, 13, 2247− 2253. (17) Kong, L.; Ziegler, G. R. Carbohydr. Polym. 2013, 92, 1416− 1422. (18) Doshi, J.; Reneker, D. H. J. Electrost. 1995, 35, 151−160. (19) Reneker, D. H.; Chun, I. Nanotechnology 1996, 7, 216. (20) Theron, S. A.; Zussman, E.; Yarin, A. L. Polymer 2004, 45, 2017−2030. (21) Rungswang, W.; Kotaki, M.; Shimojima, T.; Kimura, G.; Sakurai, S.; Chirachanchai, S. Polymer 2011, 52, 844−853. (22) Kong, L.; Ziegler, G. R. Food Hydrocolloids 2014, 36, 20−25. H

DOI: 10.1021/acs.biomac.5b00817 Biomacromolecules XXXX, XXX, XXX−XXX