- Email: [email protected]
silk sericin fibers by coaxial electrospinning
International Journal of Biological Macromolecules 51 (2012) 980–986
Contents lists available at SciVerse ScienceDirect
International Journal of Biological Macromolecules journal homepage: www.elsevier.com/locate/ijbiomac
Preparation of regenerated silk fibroin/silk sericin fibers by coaxial electrospinning Yichun Hang, Yaopeng Zhang, Yuan Jin, Huili Shao ∗ , Xuechao Hu State Key Laboratory for Modification of Chemical Fibers and Polymer Materials, College of Materials Science and Engineering, Donghua University, Shanghai 201620, PR China
a r t i c l e
i n f o
Article history: Received 13 April 2012 Received in revised form 25 June 2012 Accepted 13 August 2012 Available online 20 August 2012 Keywords: Coaxial electrospinning Silk fibroin Silk sericin Morphology Structure
a b s t r a c t The coaxial electrospinning using the regenerated silk fibroin (SF) and silk sericin (SS) aqueous solutions as the core and shell spinning dopes, respectively, was carried out to prepare regenerated SF/SS composite fibers with components and core–shell structure similar to the natural silkworm silks. It was found from the scanning electron microscope (SEM) and transmission electron microscope (TEM) results that the core dope (SF aqueous solution) flow rate (Qc ) and the applied voltage (V) had some effects on the morphology of the composite fiber. With an increase in Qc , the diameter nonuniformity and eccentricity of the core fiber became serious, while the increasing V played an inverse role. In this work, the suitable Qc for the fiber formation with better electrospinnability was about 6 L/min, and the corresponding optimum V was 40 kV. Moreover, the results from Raman spectra analysis, modulated differential scanning calorimetry (MDSC), thermogravimetry (TG) measurement and mechanical property test showed that, compared with the pure SF fiber, the coaxially electrospun SF/SS fiber had more -sheet conformation, better thermostability and mechanical properties. This was probably because that SS played significant roles in dehydrating SF molecules and inducing the conformational transition of SF to -sheet structure. © 2012 Elsevier B.V. All rights reserved.
1. Introduction The silkworm (Bombyx mori) silk has a core–shell structure and is constituted by a pair of silk fibroin (SF) core fibers covered with silk sericin (SS). The SF is the predominant component and constitutes about 75% of the total silk weight, while the SS is a kind of hydrophilic “glue-like” protein that serves as not only a cover of the SF monofilament but also an adhesive to bind two SF monofilaments together [1,2]. As a typical fibrous protein, silkworm silk shows remarkable mechanical properties and has been used in textile industry for thousands of years. More recently, regenerated SF-based biomaterials with desirable properties have attracted considerable attentions of the researchers in the field of polymer science. They have been successfully processed into different forms including films [3], 3D porous scaffolds [4], hydrogels [5], as well as fibers via wet spinning [6] or electrospinning [7]. In order to understand the formation mechanism of silkworm silk, many works have been performed to mimic the spinning process of silkworm and ultimately fabricate the silk fibers in a regenerated way. The regenerated SF fibers were first prepared
∗ Corresponding author. Tel.: +86 21 67792953; fax: +86 21 67792955. E-mail addresses: hangyc [email protected] (Y. Hang), [email protected] (Y. Zhang), [email protected] (Y. Jin), [email protected] (H. Shao), [email protected] (X. Hu). 0141-8130/$ – see front matter © 2012 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.ijbiomac.2012.08.010
through wet spinning process, usually by using polar or organic solvents and extruding the spinning dope into a coagulation bath [8–10]. However, most of these solvents either severely degrade the SF molecules or are toxic for potential industrial application. The wet spinning process has the recycling problem of the solvent used in dissolving and coagulation bath, and is rather different from the dry spinning process of silkworm using water as solvent. Recently, the electrospinning process has been developed to prepare SF-based biomaterials with a submicron diameter. In most of prior studies about the electrospinning of SF solution, the SF spinning dopes were also formed by using organic solvents [11–14], or further blending with poly (ethylene oxide), chitosan, collagen and other synthetic or nature polymers [15–17]. Only few researches were implemented using water as solvent and pure SF aqueous solution as spinning dope [18–20]. No composite spinning process using SF aqueous solution as core dope and SS aqueous solution as shell dope has been reported. It is well accepted that SF is an ideal biomaterial for tissue engineering due to its impressive mechanical properties, biocompatibility and biodegradability [21]. On the other hand, scientific investigations have also revealed that SS is resistant to oxidation and UV radiation, anti-bacterial, biocompatible, and can absorb and release moisture easily. Moreover, it exhibits many biological activities, such as tyrosinase activity inhibition, anticoagulation function, anti-cancer activity, promoting digestion and so on [22–25]. However, with weak structural properties and high solubility, pure SS biomaterials are usually fragile and difficult to
Y. Hang et al. / International Journal of Biological Macromolecules 51 (2012) 980–986
Fig. 1. Spinneret illustration for coaxial electrospinning.
fabricate [26]. In the present work, a coaxial electrospinning technique was utilized to fabricate the regenerated SF/SS fibers, which may not only mimic the compositions and core–shell structure of silkworm silk, but also achieve the mutual complementarity of SF and SS. During the electrospinning, the regenerated SF aqueous solution and SS aqueous solution were used as the core and shell spinning dope, respectively. The effects of the processing parameters, such as the core flow rate (Qc ) for SF aqueous solution and applied voltage (V), on the formation of core–shell fibers were investigated. Then, the secondary structure of the resultant electrospun SF/SS fibers were further characterized by Raman Spectroscopy and compared with the pure SF fibers. Moreover, their thermal and mechanical properties were also investigated. It is predictable that the regenerated SF/SS fibers will have great potential in the tissue engineering applications. 2. Experimental 2.1. Preparation of spinning dopes B. mori cocoons were degummed twice in boiling Na2 CO3 (0.5 wt%) aqueous solution for 30 min each at a bath ratio of 1:50 (w/v), then thoroughly rinsed to extract sericin and dried at room temperature to get degummed natural silk. The degummed silk was dissolved in 9.0 M LiBr aqueous solution at a ratio of 1:10 (w/v) at 40 ◦ C for 2 h. After being diluted, centrifugalized and then filtered, the resultant regenerated SF solution was dialyzed in deionized water at 10 ◦ C for about 3 days using the cellulose semi-permeable membrane (MWCO: 14,000 ± 2000). Finally, a 33 wt% regenerated SF aqueous solution was obtained after being subsequently condensed and used as the core spinning dope. Meanwhile, a 60 wt% regenerated SS aqueous solution was prepared by directly dissolving the SS powder (provided by Wuxi Smiss Technology Co., China) into the deionized water under gentle stirring and used as the shell spinning dope. It is known that the molecular weights (MWs) of the heavy chain and the light chain of natural silk fibroin in B. mori cocoons are 350 kDa and 26 kDa, respectively [27]. In this work, during the preparation of the spinning dope, the SF molecules were inevitably degraded due to the scission of the chains. The MW of the resultant SF after dissolution was at a range from 40 kDa to a value over 97 kDa measured by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) method as reported previously by us [28]. In addition, the MW of the water-soluble SS powder was detected to be around 2 kDa using a matrix-assisted laser desorption ionization mass spectrum (MALDI-MS, Applied Biosystems Co., USA) with a 4800 proteomics analyzer. 2.2. Electrospinning process Fig. 1 shows the spinneret illustration for coaxial electrospinning. The core (SF) and shell (SS) spinning dope were extruded
981
simultaneously through the coaxial spinneret by two separate syringe pumps. The inner tube of the spinneret has an inner diameter of 0.45 mm and an outer diameter of 0.80 mm, while the outer tube has an inner diameter of 2.00 mm and an outer diameter of 3.20 mm. Coaxial electrospinning was performed at a core flow rate (Qc ) ranging from 2 to 8 L/min and at a constant shell flow rate (Qs ) of 2 L/min. The applied voltage (V) was changed from 30 to 50 kV and the distance from the spinneret to collector was fixed to be 10 cm. The coaxially electrospun SF/SS (core/shell) fibers were collected on a grounded aluminum foil. Several copper meshes coated with carbon were previously placed on the aluminum foils to collect the fiber samples for transmission electron microscope (TEM) analysis. In this work, the electrospun pure SF fibers were also prepared as described previously for comparison [29].
2.3. Characterization The morphology of coaxial regenerated SF/SS fibers was observed using a JSM-5600LV scanning electron microscope (SEM, JEOL Co., Japan) at 15 kV. The average value (AV) and standard deviation (STDV) for the diameter of the fibers were calculated by measuring 100 individual fibers shown in the SEM images. The core–shell structures of the coaxial regenerated SF/SS fibers were investigated using an H-800 transmission electron microscope (TEM, Hitachi Co., Japan) with an accelerating voltage of 200 kV. The fiber samples deposited onto copper meshes were directly obtained by coaxial electrospinning of SF and SS dopes (see Section 2.2). Raman spectra of the electrospun fibers were obtained using a LabRam-1B microscopy Raman spectrometer (Dilor Co., France). The 632.81 nm line of a He–Ne laser was used to generate an intensity of 6 mW on the samples and the spectra were recorded from 900 to 1800 cm−1 . The quantitative analysis of the amide I region was conducted under a deconvolution method reported by Zhou et al. [30]. The three main bands of secondary structures for SF molecules are commonly employed: 1670 ± 5 cm−1 is assigned to -sheet conformation, 1655 ± 5 cm−1 to random coil/␣-helix conformation, and 1680 ± 5 cm−1 to intermediate conformation (a conformation between random coil and -sheet [30], or distorted -sheet conformation [31]). In addition, the percentage of peak area of 1615 cm−1 band assigned to the phenyl group of tyrosine residues was used as an invariant internal standard to check the validity of each analysis. In our work, this value was controlled to be about 5% of the total peak area for the above four bands. The thermal properties of the electrospun fibers were investigated using a MDSC 2910 modulated differential scanning calorimeter (TA Instruments Co., USA) under nitrogen atmosphere at a flow rate of 40 mL/min. The electrospun fiber mats were heated from room temperature to 350 ◦ C at a heating rate of 5 ◦ C/min with a modulation period of 60 s and temperature amplitude of ±1 ◦ C. Meanwhile, the thermogravimetry (TG) measurements for the electrospun fibers were also performed from room temperature to 500 ◦ C at a heating rate of 5 ◦ C/min under nitrogen atmosphere (10 mL/min) using a TG209 F1 Iris thermogravimetric apparatus (Perkin Elmer Co., USA). The kinetic evaluation of the thermal degradation of the fibers was then conducted according to the method discussed by Jimenez and Li et al. [32–34]. The mechanical properties of the electrospun fiber mats (5 mm × 40 mm) with certain thickness were investigated using an Instron 5565 material testing instrument (Instron Co., USA) at 25 ◦ C and 65 ± 5% RH. Tensile tests were performed at an extension rate of 1 mm/min, with a gauge length of 20 mm. The thicknesses of the samples were measured by using a CH-1-S thickness instrument (Shanghai Liuling Instruments Co., China) with a resolution of 0.001 mm.
982
Y. Hang et al. / International Journal of Biological Macromolecules 51 (2012) 980–986
Fig. 2. SEM images and diameter distribution histograms of the coaxially electrospun SF/SS fibers prepared at V = 40 kV, Qs = 2 L/min and under different Qc s: (a) 2 L/min, (b) 4 L/min, (c) 6 L/min and (d) 8 L/min.
3. Results and discussion 3.1. Morphology of coaxially electrospun SF/SS fibers In coaxial electrospinning, the flow rates of the spinning dopes have a significant effect on the formation of the core–shell structure of the electrospun fibers. In this work, the effect of core dope (SF solution) flow rate (Qc ) on the diameter and core–shell structure of the coaxially electrospun SF/SS fibers were investigated. Fig. 2 indicates the relationship between Qc and the morphology of coaxially electrospun SF/SS fibers. From the SEM images and diameter distribution histograms, it could be noted that the average fiber diameter (AV) almost remained unchanged (about 1600–1700 nm) despite the increase of Qc from 2 to 8 L/min. The similar result was also found in some researchers’ works on coaxially electrospun SF/poly (ethylene oxide) (PEO) or poly (DL-lactic acid) (PDLLA)/poly (3hydroxy butyrate) (PHB) core–shell fibers [35,36]. This is because that as the Qc increases, the Taylor cone becomes bigger, but the straight jet length is reduced, which results in a more significant electric field for the jet whipping process and does not lead to a remarkable increase in the diameter of the resultant electrospun fibers [36,37]. The above coaxially electrospun SF/SS fibers were further characterized by TEM to reveal the effect of Qc on the core–shell
structure and inner core diameter of the fibers. The results are shown in Fig. 3. It could be found that all the composite fibers had an obvious core–shell structure. This related to the fast stretching and solidification process of the coaxially electrospun SF/SS fibers. During the coaxial electrospinning, the charged fluid jet experiences a whipping instability after being ejected from the apex of the Taylor cone. The travel time of the jet in the air, typically on the order of milliseconds, is much shorter than the time needed for the mutual diffusion between core and shell fluids. Therefore, significant mixing between these two fluids during the coaxial electrospinning is unlikely and the sharp boundary between the core layer and shell layer can exist [38]. Although the average diameter of the resultant composite fibers exhibited a Qc -independent trend, it was further found from Fig. 3 that with a quadrupled increase in Qc , the inner core fiber diameter had an obvious increase. Meanwhile, the diameter nonuniformity and eccentricity of the core fiber became more serious. Combined with the above SEM results, it seems that the core dope (SF solution) flow rate of 6 L/min was a suitable Qc for the fabrication of the coaxial SF/SS fibers in this work. At this Qc , the resultant electrospun fibers had a relatively smaller average fiber diameter and diameter deviation, as well as a clear core–shell structure. Moreover, a thicker core of regenerated SF obtained at Qc = 6 L/min is helpful to some extent for the improvement of mechanical property of the coaxially electrospun
Fig. 3. TEM images of the coaxially electrospun SF/SS fibers prepared at V = 40 kV, Qs = 2 L/min and under different Qc s: (a) 2 L/min, (b) 4 L/min, (c) 6 L/min and (d) 8 L/min.
Y. Hang et al. / International Journal of Biological Macromolecules 51 (2012) 980–986
983
Fig. 5. Raman spectra of (a) coaxially electrospun SF/SS fiber, (b) electrospun pure SF fiber and (c) SS powder.
in coaxial jet and thus will result in the air discharge and sequentially temperature increase of the coaxial jet at the apex of spinneret [41]. In this case, the evaporation of solvent (i.e. H2 O in this work) was accelerated, and the spinneret would be obstructed easily due to the rapid solidification of SF/SS dope. As a result, the coaxial electrospinnability became relatively poor. In this work, when the core flow rate was 6 L/min, the optimum applied voltage was selected to be about 40 kV, at which a stable compound Taylor cone was formed and the coaxial SF/SS fibers with better electrospinnability and better morphology could be prepared. 3.2. Secondary structure of coaxially electrospun SF/SS fiber
Fig. 4. SEM and TEM images of the coaxially electrospun SF/SS fibers prepared at Qc = 6 L/min and different applied voltages: (a) 30 kV, (b) 40 kV and (c) 50 kV.
SF/SS fiber mats, since SF component has a predominant contribution to the mechanical property of the silk fiber [39]. Therefore, the core dope (SF solution) flow rate was maintained at 6 L/min in the further study. The effect of applied voltage (V) on the fiber diameters and morphology was also investigated. Coaxially electrospun SF/SS fibers were prepared with an increasing V from 30 to 50 kV. Fig. 4 shows the SEM and TEM images of the resultant fibers. It was observed that with the increase of V, both the average fiber diameters and the diameter deviations of the coaxial SF/SS fibers decreased, which was resulted from the increase of the strength of electric field for the jet whipping process. Moreover, the TEM images further showed that the SF/SS fibers with core–shell structure could be successfully obtained using coaxial electrospinning under each V, but differed in the uniformity of the core and shell layers. When V was not sufficiently high (e.g. 30 kV), the coaxial jet charged by a lower electric field would drip from the apex of spinneret because of the gravitation [40]. This resulted in the inefficient and nonsimultaneous stretching of core and shell spinning dopes, and consequently the formation of fibers with a larger diameter and nonuniform core–shell structure as shown in Fig. 4(a). When V was increased to 40 kV, the average fiber diameter decreased. However, sequentially increasing V to 50 kV caused the unstableness of Taylor cone, although it induced the further decreasing of fiber diameter. The high voltage is very apt to lead to the excess of electrostatic charge
Fig. 5 shows the Raman spectrum of the coaxially electrospun SF/SS fiber. The Raman results of the electrospun pure SF fiber and SS powder were also shown in the figure for comparison. It was found that both electrospun fibers exhibited characteristic bands at 1666 cm−1 (in the amide I region), 1254 cm−1 (in the amide III region) and 1107 cm−1 (in the C C stretch region), which were attributed to the -sheet, random coil and ␣-helix conformation, respectively [42,43]. In addition, for the coaxially electrospun SF/SS fiber, the characteristic bands of random coil conformation for SS at 1652 cm−1 and 1275 cm−1 were also exhibited, which could be observed clearly in the Raman result of SS powder. Moreover, the SF/SS fiber had another two characteristic bands at 1085 cm−1 and 1680 cm−1 . The former was assigned to the -sheet conformation [44,45] and the latter was assigned to the intermediate conformation related to the -sheet. By contrast, for the pure SF fiber, these two bands were exhibited weakly as shoulder peaks. The above results suggested that the coaxially electrospun SF/SS fiber had more stable structure with more -sheet (and related intermediate) conformation compared with the electrospun pure SF fiber. A deconvolution method (see Section 2.3) was further used to conduct the quantitative structural analysis from the amide I region of Raman spectra for the electrospun fibers. Using this method, we obtained the estimated quantities of an invariant internal standard (phenyl) and three main components of secondary structure for SF. The results were listed in Table 1. It was also confirmed that with the covering of SS, the composition of random coil/␣-helix conformation became lower, while that of the -sheet conformation and the intermediate conformation related to -sheet became higher. This indicated that the presence of SS facilitated the conformational transition of SF from random coil/␣-helix structure to -sheet (and/or related intermediate) conformation. Its possible mechanism was schematically shown in Fig. 6. Since the typical amino acid sequence of repetitive motif for SS is a characteristic 38-amino
984
Y. Hang et al. / International Journal of Biological Macromolecules 51 (2012) 980–986
Table 1 Secondary structural compositions of regenerated silk fibers by electrospinning. Fiber sample
a. Coaxial SF/SS b. SF
Contents of four components estimated from amide I region (%) Phenyl group
Random coil/␣-helix
-Sheet
Intermediate conformation
4.9 5.1
26.3 38.3
29.5 23.4
39.3 33.2
acid peptide rich in Ser [46], SS has a higher content (up to 63%) of hydrophilic amino acid side chains (depicted as white character on a black background in Fig. 6) than SF (about 16%) [27,47]. Thus, it can absorb more water molecules surrounding SF and dehydrate
SF through the interfacial interaction during coaxial electrospinning (see Fig. 6(b)). This gives rise to the regular arrangement of SF chains by forming hydrogen bonds between SF molecules and consequently the formation of -sheet conformation (as demonstrated
Fig. 6. Schematic diagram of the conformational transition mechanism of SF during the coaxial electrospinning of SF and SS aqueous solutions. The typical amino acid sequence of repetitive motif for the SF is GAGAGS domain [27], while for SS is a characteristic 38-amino acid peptide rich in Ser (SVSSTGSSSNTDSNSNSAGSSTSGGSSTYGYSSNSRDG) [46]. The amino acids with hydrophilic side chains in SF and SS molecules are depicted as white character on a black background.
Y. Hang et al. / International Journal of Biological Macromolecules 51 (2012) 980–986
985
Fig. 7. Non-reversing (upper curve, i) and reversing (lower curve, ii) heat flow results of the coaxially electrospun SF/SS silk fiber (A), and the total heat flow curves for the coaxial SF/SS and pure SF fibers prepared by electrospinning (B).
in Fig. 6(b and c)). Therefore, it is hypothesized from the above analysis that besides serving as an adhesive to bind two SF monofilaments together in cocoon, SS may also play a role in accelerating the reduction of the water content surrounding SF and inducing the conformational transition of SF to -sheet structure.
3.3. Thermal properties of coaxially electrospun SF/SS fiber The thermal properties of coaxially electrospun SF/SS fiber were further investigated by MDSC analysis. The non-reversing (upper curve, i) and reversing (lower curve, ii) heat flow curves of the fiber ranging from 100 to 350 ◦ C were shown in Fig. 7(A). The wide endothermic peak centered at about 125 ◦ C in the non-reversing heat flow curve was due to the heat absorbed by the evaporated water molecules as well as the heat of transformation of the bound water-silk structure to the dry silk structure [48]. The endothermic shift of the baseline without any sharp transition peaks at around 180 ◦ C in the reversing heat flow curve was assigned to the glass transition of silk fibroin [48–51]. Similar shift corresponding to glass transition was also found in the DSC curves of thermotropic liquid crystalline copolyesters [52,53]. Above the temperature of glass transition (Tg ), the non-reversing heat flow curve exhibited a large exothermic peak centered at about 278 ◦ C, which was attributed to the non-isothermal crystallization from the secondary structure of noncrystalline random coil or ␣-helix to -sheet crystal. After the appearance of the crystallization peak, the coaxially electrospun SF/SS fiber exhibited a large endothermic peak at around 290 ◦ C, indicating that the resultant fiber started to degrade [54,55]. Furthermore, Fig. 7(B) showed the total heat flow curves for pure SF fiber and the coaxial SF/SS fiber prepared by electrospinning. It was worth noting that the coaxial SF/SS fiber had a higher degradation temperature (290 ◦ C) than the pure SF fiber (282 ◦ C), which coincided with the results from the TG analysis. According to the method discussed by Jimenez and Li et al. [32–34], some thermal degradation data and kinetic parameters of degradation calculated from the TG results for these two fibers were obtained and shown in Table 2. It was found that, for the coaxial SF/SS fiber, not only its degradation temperature at the maximum weight-loss rate (Tdm ) and residue mass at 500 ◦ C, but also the activation energy (E), the frequency factor (ln Z) and the order (n) of its thermal degradation reaction were all higher than those for the pure SF fiber. On the other hand, the former had a lower maximum weight-loss rate than the latter. These results indicated that the coaxial SF/SS fiber was more thermally stable than the pure SF fiber, which can be attributed to the more stable structure with more -sheet conformation in the coaxial SF/SS fiber.
Table 2 Thermal degradation data and kinetic parameters of degradation calculated from the TG results of regenerated silk fiber mats by electrospinning. Parameters
Fiber sample
Tdm (◦ C) Maximum weight-loss rate (%/min) Residue mass at 500 ◦ C (%) E (kJ/mol) ln Z (1/min) n
Coaxial SF/SS
SF
293 1.86 40.01 45.69 11.66 3.85
286 1.91 38.52 36.87 9.67 3.09
In addition, Fig. 7(B) also displayed the temperature difference in water evaporation of the above two fibers and revealed that the coaxial SF/SS fiber had a lower water evaporation temperature (125 ◦ C) than SF fiber (146 ◦ C). The difference in water evaporation temperature between these two fibers may be concerned with the bound water molecules in the fibers which were relatively hard to be evaporated [48]. It was thought that, for the coaxial SF/SS fiber, the dehydration effect taken by SS considerably decreased the content of the bound water molecules surrounding SF and enabled the transformation of some bound water molecules to free water. This resulted in an easier water loss process and lower temperature for water evaporation in coaxially electrospun SF/SS fiber. 3.4. Mechanical properties of coaxially electrospun SF/SS fiber The mechanical properties of the coaxial SF/SS and pure SF fiber mats prepared by electrospinning were shown in Fig. 8 and Table 3. It was found that the mechanical properties of the electrospun fiber mats were affected obviously by the covering of SS. Compared with the SF fiber mat, the average breaking strength and breaking energy of coaxial SF/SS fiber mat were increased by nearly 82% (to 1.93 MPa) and 93% (to 7.21 J/kg), respectively. One of the main reasons for the improvement of the mechanical properties of the SF/SS fiber mat is the contribution of SS in inducing more -sheet conformation of the SF molecules in the coaxially electrospun SF/SS fiber.
Table 3 Mechanical properties of regenerated silk fiber mats by electrospinning. Fiber sample
Breaking elongation (%)
Breaking strength (MPa)
Breaking energy (J/kg)
a. Coaxial SF/SS b. SF
1.20 ± 0.55 0.96 ± 0.16
1.93 ± 0.40 1.06 ± 0.23
7.21 ± 4.14 3.74 ± 1.75
986
Y. Hang et al. / International Journal of Biological Macromolecules 51 (2012) 980–986
Fig. 8. Stress–strain curves of the regenerated silk fiber mats by electrospinning: (a) coaxial SF/SS and (b) SF.
4. Conclusions This work has demonstrated the feasibility of utilizing the coaxial electrospinning technique to prepare regenerated silk fibroin/silk sericin (SF/SS) fibers in order to mimic the compositions and core–shell structure of the silkworm silk. Based on the morphological analysis of the coaxially electrospun SF/SS fibers, we have acquired useful understanding about the effects of several processing parameters on the formation of core–shell fibers during the coaxial electrospinning. Investigation into the effect of the core dope (SF solution) flow rate (Qc ) revealed that the average fiber diameter of the composite fibers exhibited a Qc -independent trend. However, with a quadrupled increase in Qc , the inner core fiber diameter had an obvious increase and the diameter nonuniformity and eccentricity of the core fiber became more serious. In this work, the Qc of 6 L/min was suitable for the fabrication of coaxial SF/SS fibers with core–shell structure and better electrospinnability, and the corresponding optimum applied voltage (V) was about 40 kV. Moreover, compared with the electrospun pure SF fibers, the coaxial SF/SS fibers had more stable structure with more -sheet (and related intermediate) conformation and accordingly better thermostability and mechanical properties, since SS had a higher content of hydrophilic amino acid side chains and played significant roles in dehydrating SF molecules and inducing the conformational transition of SF to -sheet structure. Acknowledgments This work is supported by the National Natural Science Foundation of China (81170641), Specialized Research Fund for the Doctoral Program of Higher Education (200802550001), Innovation Program of Shanghai Municipal Education Commission (12ZZ065), Shanghai Rising-Star Program (12QA1400100) and the Fundamental Research Funds for the Central Universities. References [1] Y. Shen, M.A. Johnson, D.C. Martin, Macromolecules 31 (1998) 8857–8864. [2] A. Matsumoto, J. Chen, A.L. Collette, U.J. Kim, G.H. Altman, P. Cebe, D.L. Kaplan, Journal of Physical Chemistry B 110 (2006) 21630–21638. [3] H.J. Jin, J. Park, V. Karageorgiou, U.J. Kim, R. Valluzzi, D.L. Kaplan, Advanced Functional Materials 15 (2005) 1241–1247. [4] U.J. Kim, J. Park, H.J. Kim, M. Wada, D.L. Kaplan, Biomaterials 26 (2005) 2775–2785. [5] U.J. Kim, J.Y. Park, C.M. Li, H.J. Jin, R. Valluzzi, D.L. Kaplan, Biomacromolecules 5 (2004) 786–792. [6] J.P. Yan, G.Q. Zhou, D.P. Knight, Z.Z. Shao, X. Chen, Biomacromolecules 11 (2010) 1–5.
[7] J.X. He, Y.R. Qin, S.Z. Cui, Y.Y. Gao, S.Y. Wang, Journal of Materials Science 46 (2011) 2938–2946. [8] K.A. Trabbic, P. Yager, Macromolecules 31 (1998) 462–471. [9] K.H. Lee, D.H. Baek, C.S. Ki, Y.H. Park, International Journal of Biological Macromolecules 41 (2007) 168–172. [10] E. Marsano, P. Corsini, C. Arosio, A. Boschi, M. Mormino, G. Freddi, International Journal of Biological Macromolecules 37 (2005) 179–188. [11] S. Zarkoob, D.H. Reneker, R.K. Eby, S.D. Hudson, D. Ertley, W.W. Adams, Polymer Preprints 39 (1998) 244–245. [12] J. Ayutsede, M. Gandhi, S. Sukigara, M. Micklus, H.E. Chen, F. Ko, Polymer 46 (2005) 1625–1634. [13] Y. Kawahara, A. Nakayama, N. Matsumura, T. Yoshioka, M. Tsuji, Journal of Applied Polymer Science 107 (2008) 3681–3684. [14] F. Zhang, B. Zuo, Z. Fan, Z. Xie, Q. Lu, X. Zhang, D.L. Kaplan, Biomacromolecules 13 (2012) 798–804. [15] H.J. Jin, S.V. Fridrikh, G.C. Rutledge, D.L. Kaplan, Biomacromolecules 3 (2002) 1233–1239. [16] W.H. Park, L. Jeong, D.I. Yoo, S. Hudson, Polymer 45 (2004) 7151–7157. [17] S. Putthanarat, R.K. Eby, W. Kataphinan, S. Jones, R. Naik, D.H. Reneker, B.L. Farmer, Polymer 47 (2006) 5630–5632. [18] H. Wang, Y.P. Zhang, H.L. Shao, X.C. Hu, Journal of Materials Science 40 (2005) 5359–5363. [19] J.X. Zhu, Y.P. Zhang, H.L. Shao, X.C. Hu, Polymer 49 (2008) 2880–2885. [20] H. Cao, X. Chen, L. Huang, Z.Z. Shao, Materials Science and Engineering C 29 (2009) 2270–2274. [21] G.H. Altman, F. Diaz, C. Jakuba, T. Calabro, R.L. Horan, J. Chen, H. Lu, J. Richmond, D.L. Kaplan, Biomaterials 24 (2003) 401–416. [22] N. Kato, S. Sato, A. Yamanaka, H. Yamada, N. Fuwa, M. Nomura, Bioscience, Biotechnology, and Biochemistry 62 (1998) 145–147. [23] X. Zhang, M.M.R. Khan, T. Yamamoto, M. Tsukada, H. Morikawa, International Journal of Biological Macromolecules 50 (2012) 337–347. [24] M. Sasaki, H. Yamada, N. Kato, Nutrition Research 20 (2000) 1505–1511. [25] S. Zhaorigetu, M. Sasaki, H. Watanabe, N. Kato, Bioscience, Biotechnology and Biochemistry 65 (2001) 2181–2186. [26] M.L. Gimenes, L. Liu, X. Feng, Journal of Membrane Science 295 (2007) 71–79. [27] C.J. Fu, Z.Z. Shao, V. Fritz, Chemical Communications 43 (2009) 6515–6529. [28] W. Wei, Y.P. Zhang, H.L. Shao, X.C. Hu, Journal of AOAC International 93 (2010) 1143–1147. [29] Y.C. Hang, Y.P. Zhang, Y. Jin, H.L. Shao, X.C. Hu, Journal of Materials Research 26 (2011) 2931–2937. [30] L. Zhou, X. Chen, Z.Z. Shao, Y.F. Huang, D.P. Knight, Journal of Physical Chemistry B 109 (2005) 16937–16945. [31] P. Zhou, X. Xie, D.P. Knight, X.H. Zong, F. Deng, W.H. Yao, Biochemistry 43 (2004) 11302–11311. [32] A. Jimenez, V. Berenguer, J. Lopez, A. Sanchez, Journal of Applied Polymer Science 50 (1993) 1565–1573. [33] X.G. Li, M.R. Huang, Polymer Degradation and Stability 64 (1999) 81–90. [34] X.G. Li, M.R. Huang, Y.Q. Lu, M.F. Zhu, Journal of Materials Chemistry 15 (2005) 1343–1352. [35] M. Wang, J.H. Yu, D.L. Kaplan, G.C. Rutledge, Macromolecules 39 (2006) 1102–1107. [36] C. Wang, K.W. Yan, Y.D. Lin, P.C.H. Hsieh, Macromolecules 43 (2010) 6389–6397. [37] C. Wang, Y.W. Cheng, C.H. Hsu, H.S. Chien, S.Y. Tsou, Journal of Polymer Research 18 (2011) 111–123. [38] Z.C. Sun, E. Zussman, A.L. Yarin, J.H. Wendorff, A. Greiner, Advanced Materials 15 (2003) 1929–1932. [39] J. Perez-Rigueiro, M. Elices, J. Llorca, C. Viney, Journal of Applied Polymer Science 82 (2001) 1928–1935. [40] A.K. Moghe, B.S. Gupta, Polymer Reviews 48 (2008) 353–377. [41] J.M. Deitzel, J. Kleinmeyer, D. Harris, N.C. Beck Tan, Polymer 42 (2001) 261–272. [42] P. Monti, G. Freddi, A. Bertoluzza, N. Kasai, M. Tsukada, Journal of Raman Spectroscopy 29 (1998) 297–304. [43] P. Monti, P. Taddei, G. Freddi, T. Asakura, M. Tsukada, Journal of Raman Spectroscopy 32 (2001) 103–107. [44] X.G. Li, L.Y. Wu, M.R. Huang, H.L. Shao, X.C. Hu, Biopolymers 89 (2008) 497–505. [45] R. Tuma, P.E. Prevelige, G.J. Thomas, Biochemistry 35 (1996) 4619–4627. [46] H. Teramoto, A. Kakazu, K. Yamauchi, T. Asakura, Macromolecules 40 (2007) 1562–1569. [47] H. Teramoto, T. Kameda, Y. Tamada, Bioscience, Biotechnology, and Biochemistry 72 (2008) 3189–3196. [48] X. Hu, D. Kaplan, P. Cebe, Macromolecules 41 (2008) 3939–3948. [49] X. Hu, D. Kaplan, P. Cebe, Thermochimica Acta 461 (2007) 137–144. [50] A. Motta, L. Fambri, C. Migliaresi, Macromolecular Chemistry and Physics 203 (2002) 1658–1665. [51] N. Agarwal, D.A. Hoagland, R.J. Farris, Journal of Applied Polymer Science 63 (1997) 401–410. [52] X.G. Li, Z.L. Zhou, X.G. Wu, T. Sun, Journal of Applied Polymer Science 51 (1994) 1913–1921. [53] X.G. Li, M.R. Huang, G.H. Guan, T. Sun, Journal of Applied Polymer Science 66 (1997) 2129–2138. [54] G. Freddi, G. Pessina, M. Tsukada, International Journal of Biological Macromolecules 24 (1999) 251–263. [55] Y. Tao, W. Xu, Y. Yan, Y. Cao, Polymer International 61 (2012) 760–767.
Contents lists available at SciVerse ScienceDirect
International Journal of Biological Macromolecules journal homepage: www.elsevier.com/locate/ijbiomac
Preparation of regenerated silk fibroin/silk sericin fibers by coaxial electrospinning Yichun Hang, Yaopeng Zhang, Yuan Jin, Huili Shao ∗ , Xuechao Hu State Key Laboratory for Modification of Chemical Fibers and Polymer Materials, College of Materials Science and Engineering, Donghua University, Shanghai 201620, PR China
a r t i c l e
i n f o
Article history: Received 13 April 2012 Received in revised form 25 June 2012 Accepted 13 August 2012 Available online 20 August 2012 Keywords: Coaxial electrospinning Silk fibroin Silk sericin Morphology Structure
a b s t r a c t The coaxial electrospinning using the regenerated silk fibroin (SF) and silk sericin (SS) aqueous solutions as the core and shell spinning dopes, respectively, was carried out to prepare regenerated SF/SS composite fibers with components and core–shell structure similar to the natural silkworm silks. It was found from the scanning electron microscope (SEM) and transmission electron microscope (TEM) results that the core dope (SF aqueous solution) flow rate (Qc ) and the applied voltage (V) had some effects on the morphology of the composite fiber. With an increase in Qc , the diameter nonuniformity and eccentricity of the core fiber became serious, while the increasing V played an inverse role. In this work, the suitable Qc for the fiber formation with better electrospinnability was about 6 L/min, and the corresponding optimum V was 40 kV. Moreover, the results from Raman spectra analysis, modulated differential scanning calorimetry (MDSC), thermogravimetry (TG) measurement and mechanical property test showed that, compared with the pure SF fiber, the coaxially electrospun SF/SS fiber had more -sheet conformation, better thermostability and mechanical properties. This was probably because that SS played significant roles in dehydrating SF molecules and inducing the conformational transition of SF to -sheet structure. © 2012 Elsevier B.V. All rights reserved.
1. Introduction The silkworm (Bombyx mori) silk has a core–shell structure and is constituted by a pair of silk fibroin (SF) core fibers covered with silk sericin (SS). The SF is the predominant component and constitutes about 75% of the total silk weight, while the SS is a kind of hydrophilic “glue-like” protein that serves as not only a cover of the SF monofilament but also an adhesive to bind two SF monofilaments together [1,2]. As a typical fibrous protein, silkworm silk shows remarkable mechanical properties and has been used in textile industry for thousands of years. More recently, regenerated SF-based biomaterials with desirable properties have attracted considerable attentions of the researchers in the field of polymer science. They have been successfully processed into different forms including films [3], 3D porous scaffolds [4], hydrogels [5], as well as fibers via wet spinning [6] or electrospinning [7]. In order to understand the formation mechanism of silkworm silk, many works have been performed to mimic the spinning process of silkworm and ultimately fabricate the silk fibers in a regenerated way. The regenerated SF fibers were first prepared
∗ Corresponding author. Tel.: +86 21 67792953; fax: +86 21 67792955. E-mail addresses: hangyc [email protected] (Y. Hang), [email protected] (Y. Zhang), [email protected] (Y. Jin), [email protected] (H. Shao), [email protected] (X. Hu). 0141-8130/$ – see front matter © 2012 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.ijbiomac.2012.08.010
through wet spinning process, usually by using polar or organic solvents and extruding the spinning dope into a coagulation bath [8–10]. However, most of these solvents either severely degrade the SF molecules or are toxic for potential industrial application. The wet spinning process has the recycling problem of the solvent used in dissolving and coagulation bath, and is rather different from the dry spinning process of silkworm using water as solvent. Recently, the electrospinning process has been developed to prepare SF-based biomaterials with a submicron diameter. In most of prior studies about the electrospinning of SF solution, the SF spinning dopes were also formed by using organic solvents [11–14], or further blending with poly (ethylene oxide), chitosan, collagen and other synthetic or nature polymers [15–17]. Only few researches were implemented using water as solvent and pure SF aqueous solution as spinning dope [18–20]. No composite spinning process using SF aqueous solution as core dope and SS aqueous solution as shell dope has been reported. It is well accepted that SF is an ideal biomaterial for tissue engineering due to its impressive mechanical properties, biocompatibility and biodegradability [21]. On the other hand, scientific investigations have also revealed that SS is resistant to oxidation and UV radiation, anti-bacterial, biocompatible, and can absorb and release moisture easily. Moreover, it exhibits many biological activities, such as tyrosinase activity inhibition, anticoagulation function, anti-cancer activity, promoting digestion and so on [22–25]. However, with weak structural properties and high solubility, pure SS biomaterials are usually fragile and difficult to
Y. Hang et al. / International Journal of Biological Macromolecules 51 (2012) 980–986
Fig. 1. Spinneret illustration for coaxial electrospinning.
fabricate [26]. In the present work, a coaxial electrospinning technique was utilized to fabricate the regenerated SF/SS fibers, which may not only mimic the compositions and core–shell structure of silkworm silk, but also achieve the mutual complementarity of SF and SS. During the electrospinning, the regenerated SF aqueous solution and SS aqueous solution were used as the core and shell spinning dope, respectively. The effects of the processing parameters, such as the core flow rate (Qc ) for SF aqueous solution and applied voltage (V), on the formation of core–shell fibers were investigated. Then, the secondary structure of the resultant electrospun SF/SS fibers were further characterized by Raman Spectroscopy and compared with the pure SF fibers. Moreover, their thermal and mechanical properties were also investigated. It is predictable that the regenerated SF/SS fibers will have great potential in the tissue engineering applications. 2. Experimental 2.1. Preparation of spinning dopes B. mori cocoons were degummed twice in boiling Na2 CO3 (0.5 wt%) aqueous solution for 30 min each at a bath ratio of 1:50 (w/v), then thoroughly rinsed to extract sericin and dried at room temperature to get degummed natural silk. The degummed silk was dissolved in 9.0 M LiBr aqueous solution at a ratio of 1:10 (w/v) at 40 ◦ C for 2 h. After being diluted, centrifugalized and then filtered, the resultant regenerated SF solution was dialyzed in deionized water at 10 ◦ C for about 3 days using the cellulose semi-permeable membrane (MWCO: 14,000 ± 2000). Finally, a 33 wt% regenerated SF aqueous solution was obtained after being subsequently condensed and used as the core spinning dope. Meanwhile, a 60 wt% regenerated SS aqueous solution was prepared by directly dissolving the SS powder (provided by Wuxi Smiss Technology Co., China) into the deionized water under gentle stirring and used as the shell spinning dope. It is known that the molecular weights (MWs) of the heavy chain and the light chain of natural silk fibroin in B. mori cocoons are 350 kDa and 26 kDa, respectively [27]. In this work, during the preparation of the spinning dope, the SF molecules were inevitably degraded due to the scission of the chains. The MW of the resultant SF after dissolution was at a range from 40 kDa to a value over 97 kDa measured by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) method as reported previously by us [28]. In addition, the MW of the water-soluble SS powder was detected to be around 2 kDa using a matrix-assisted laser desorption ionization mass spectrum (MALDI-MS, Applied Biosystems Co., USA) with a 4800 proteomics analyzer. 2.2. Electrospinning process Fig. 1 shows the spinneret illustration for coaxial electrospinning. The core (SF) and shell (SS) spinning dope were extruded
981
simultaneously through the coaxial spinneret by two separate syringe pumps. The inner tube of the spinneret has an inner diameter of 0.45 mm and an outer diameter of 0.80 mm, while the outer tube has an inner diameter of 2.00 mm and an outer diameter of 3.20 mm. Coaxial electrospinning was performed at a core flow rate (Qc ) ranging from 2 to 8 L/min and at a constant shell flow rate (Qs ) of 2 L/min. The applied voltage (V) was changed from 30 to 50 kV and the distance from the spinneret to collector was fixed to be 10 cm. The coaxially electrospun SF/SS (core/shell) fibers were collected on a grounded aluminum foil. Several copper meshes coated with carbon were previously placed on the aluminum foils to collect the fiber samples for transmission electron microscope (TEM) analysis. In this work, the electrospun pure SF fibers were also prepared as described previously for comparison [29].
2.3. Characterization The morphology of coaxial regenerated SF/SS fibers was observed using a JSM-5600LV scanning electron microscope (SEM, JEOL Co., Japan) at 15 kV. The average value (AV) and standard deviation (STDV) for the diameter of the fibers were calculated by measuring 100 individual fibers shown in the SEM images. The core–shell structures of the coaxial regenerated SF/SS fibers were investigated using an H-800 transmission electron microscope (TEM, Hitachi Co., Japan) with an accelerating voltage of 200 kV. The fiber samples deposited onto copper meshes were directly obtained by coaxial electrospinning of SF and SS dopes (see Section 2.2). Raman spectra of the electrospun fibers were obtained using a LabRam-1B microscopy Raman spectrometer (Dilor Co., France). The 632.81 nm line of a He–Ne laser was used to generate an intensity of 6 mW on the samples and the spectra were recorded from 900 to 1800 cm−1 . The quantitative analysis of the amide I region was conducted under a deconvolution method reported by Zhou et al. [30]. The three main bands of secondary structures for SF molecules are commonly employed: 1670 ± 5 cm−1 is assigned to -sheet conformation, 1655 ± 5 cm−1 to random coil/␣-helix conformation, and 1680 ± 5 cm−1 to intermediate conformation (a conformation between random coil and -sheet [30], or distorted -sheet conformation [31]). In addition, the percentage of peak area of 1615 cm−1 band assigned to the phenyl group of tyrosine residues was used as an invariant internal standard to check the validity of each analysis. In our work, this value was controlled to be about 5% of the total peak area for the above four bands. The thermal properties of the electrospun fibers were investigated using a MDSC 2910 modulated differential scanning calorimeter (TA Instruments Co., USA) under nitrogen atmosphere at a flow rate of 40 mL/min. The electrospun fiber mats were heated from room temperature to 350 ◦ C at a heating rate of 5 ◦ C/min with a modulation period of 60 s and temperature amplitude of ±1 ◦ C. Meanwhile, the thermogravimetry (TG) measurements for the electrospun fibers were also performed from room temperature to 500 ◦ C at a heating rate of 5 ◦ C/min under nitrogen atmosphere (10 mL/min) using a TG209 F1 Iris thermogravimetric apparatus (Perkin Elmer Co., USA). The kinetic evaluation of the thermal degradation of the fibers was then conducted according to the method discussed by Jimenez and Li et al. [32–34]. The mechanical properties of the electrospun fiber mats (5 mm × 40 mm) with certain thickness were investigated using an Instron 5565 material testing instrument (Instron Co., USA) at 25 ◦ C and 65 ± 5% RH. Tensile tests were performed at an extension rate of 1 mm/min, with a gauge length of 20 mm. The thicknesses of the samples were measured by using a CH-1-S thickness instrument (Shanghai Liuling Instruments Co., China) with a resolution of 0.001 mm.
982
Y. Hang et al. / International Journal of Biological Macromolecules 51 (2012) 980–986
Fig. 2. SEM images and diameter distribution histograms of the coaxially electrospun SF/SS fibers prepared at V = 40 kV, Qs = 2 L/min and under different Qc s: (a) 2 L/min, (b) 4 L/min, (c) 6 L/min and (d) 8 L/min.
3. Results and discussion 3.1. Morphology of coaxially electrospun SF/SS fibers In coaxial electrospinning, the flow rates of the spinning dopes have a significant effect on the formation of the core–shell structure of the electrospun fibers. In this work, the effect of core dope (SF solution) flow rate (Qc ) on the diameter and core–shell structure of the coaxially electrospun SF/SS fibers were investigated. Fig. 2 indicates the relationship between Qc and the morphology of coaxially electrospun SF/SS fibers. From the SEM images and diameter distribution histograms, it could be noted that the average fiber diameter (AV) almost remained unchanged (about 1600–1700 nm) despite the increase of Qc from 2 to 8 L/min. The similar result was also found in some researchers’ works on coaxially electrospun SF/poly (ethylene oxide) (PEO) or poly (DL-lactic acid) (PDLLA)/poly (3hydroxy butyrate) (PHB) core–shell fibers [35,36]. This is because that as the Qc increases, the Taylor cone becomes bigger, but the straight jet length is reduced, which results in a more significant electric field for the jet whipping process and does not lead to a remarkable increase in the diameter of the resultant electrospun fibers [36,37]. The above coaxially electrospun SF/SS fibers were further characterized by TEM to reveal the effect of Qc on the core–shell
structure and inner core diameter of the fibers. The results are shown in Fig. 3. It could be found that all the composite fibers had an obvious core–shell structure. This related to the fast stretching and solidification process of the coaxially electrospun SF/SS fibers. During the coaxial electrospinning, the charged fluid jet experiences a whipping instability after being ejected from the apex of the Taylor cone. The travel time of the jet in the air, typically on the order of milliseconds, is much shorter than the time needed for the mutual diffusion between core and shell fluids. Therefore, significant mixing between these two fluids during the coaxial electrospinning is unlikely and the sharp boundary between the core layer and shell layer can exist [38]. Although the average diameter of the resultant composite fibers exhibited a Qc -independent trend, it was further found from Fig. 3 that with a quadrupled increase in Qc , the inner core fiber diameter had an obvious increase. Meanwhile, the diameter nonuniformity and eccentricity of the core fiber became more serious. Combined with the above SEM results, it seems that the core dope (SF solution) flow rate of 6 L/min was a suitable Qc for the fabrication of the coaxial SF/SS fibers in this work. At this Qc , the resultant electrospun fibers had a relatively smaller average fiber diameter and diameter deviation, as well as a clear core–shell structure. Moreover, a thicker core of regenerated SF obtained at Qc = 6 L/min is helpful to some extent for the improvement of mechanical property of the coaxially electrospun
Fig. 3. TEM images of the coaxially electrospun SF/SS fibers prepared at V = 40 kV, Qs = 2 L/min and under different Qc s: (a) 2 L/min, (b) 4 L/min, (c) 6 L/min and (d) 8 L/min.
Y. Hang et al. / International Journal of Biological Macromolecules 51 (2012) 980–986
983
Fig. 5. Raman spectra of (a) coaxially electrospun SF/SS fiber, (b) electrospun pure SF fiber and (c) SS powder.
in coaxial jet and thus will result in the air discharge and sequentially temperature increase of the coaxial jet at the apex of spinneret [41]. In this case, the evaporation of solvent (i.e. H2 O in this work) was accelerated, and the spinneret would be obstructed easily due to the rapid solidification of SF/SS dope. As a result, the coaxial electrospinnability became relatively poor. In this work, when the core flow rate was 6 L/min, the optimum applied voltage was selected to be about 40 kV, at which a stable compound Taylor cone was formed and the coaxial SF/SS fibers with better electrospinnability and better morphology could be prepared. 3.2. Secondary structure of coaxially electrospun SF/SS fiber
Fig. 4. SEM and TEM images of the coaxially electrospun SF/SS fibers prepared at Qc = 6 L/min and different applied voltages: (a) 30 kV, (b) 40 kV and (c) 50 kV.
SF/SS fiber mats, since SF component has a predominant contribution to the mechanical property of the silk fiber [39]. Therefore, the core dope (SF solution) flow rate was maintained at 6 L/min in the further study. The effect of applied voltage (V) on the fiber diameters and morphology was also investigated. Coaxially electrospun SF/SS fibers were prepared with an increasing V from 30 to 50 kV. Fig. 4 shows the SEM and TEM images of the resultant fibers. It was observed that with the increase of V, both the average fiber diameters and the diameter deviations of the coaxial SF/SS fibers decreased, which was resulted from the increase of the strength of electric field for the jet whipping process. Moreover, the TEM images further showed that the SF/SS fibers with core–shell structure could be successfully obtained using coaxial electrospinning under each V, but differed in the uniformity of the core and shell layers. When V was not sufficiently high (e.g. 30 kV), the coaxial jet charged by a lower electric field would drip from the apex of spinneret because of the gravitation [40]. This resulted in the inefficient and nonsimultaneous stretching of core and shell spinning dopes, and consequently the formation of fibers with a larger diameter and nonuniform core–shell structure as shown in Fig. 4(a). When V was increased to 40 kV, the average fiber diameter decreased. However, sequentially increasing V to 50 kV caused the unstableness of Taylor cone, although it induced the further decreasing of fiber diameter. The high voltage is very apt to lead to the excess of electrostatic charge
Fig. 5 shows the Raman spectrum of the coaxially electrospun SF/SS fiber. The Raman results of the electrospun pure SF fiber and SS powder were also shown in the figure for comparison. It was found that both electrospun fibers exhibited characteristic bands at 1666 cm−1 (in the amide I region), 1254 cm−1 (in the amide III region) and 1107 cm−1 (in the C C stretch region), which were attributed to the -sheet, random coil and ␣-helix conformation, respectively [42,43]. In addition, for the coaxially electrospun SF/SS fiber, the characteristic bands of random coil conformation for SS at 1652 cm−1 and 1275 cm−1 were also exhibited, which could be observed clearly in the Raman result of SS powder. Moreover, the SF/SS fiber had another two characteristic bands at 1085 cm−1 and 1680 cm−1 . The former was assigned to the -sheet conformation [44,45] and the latter was assigned to the intermediate conformation related to the -sheet. By contrast, for the pure SF fiber, these two bands were exhibited weakly as shoulder peaks. The above results suggested that the coaxially electrospun SF/SS fiber had more stable structure with more -sheet (and related intermediate) conformation compared with the electrospun pure SF fiber. A deconvolution method (see Section 2.3) was further used to conduct the quantitative structural analysis from the amide I region of Raman spectra for the electrospun fibers. Using this method, we obtained the estimated quantities of an invariant internal standard (phenyl) and three main components of secondary structure for SF. The results were listed in Table 1. It was also confirmed that with the covering of SS, the composition of random coil/␣-helix conformation became lower, while that of the -sheet conformation and the intermediate conformation related to -sheet became higher. This indicated that the presence of SS facilitated the conformational transition of SF from random coil/␣-helix structure to -sheet (and/or related intermediate) conformation. Its possible mechanism was schematically shown in Fig. 6. Since the typical amino acid sequence of repetitive motif for SS is a characteristic 38-amino
984
Y. Hang et al. / International Journal of Biological Macromolecules 51 (2012) 980–986
Table 1 Secondary structural compositions of regenerated silk fibers by electrospinning. Fiber sample
a. Coaxial SF/SS b. SF
Contents of four components estimated from amide I region (%) Phenyl group
Random coil/␣-helix
-Sheet
Intermediate conformation
4.9 5.1
26.3 38.3
29.5 23.4
39.3 33.2
acid peptide rich in Ser [46], SS has a higher content (up to 63%) of hydrophilic amino acid side chains (depicted as white character on a black background in Fig. 6) than SF (about 16%) [27,47]. Thus, it can absorb more water molecules surrounding SF and dehydrate
SF through the interfacial interaction during coaxial electrospinning (see Fig. 6(b)). This gives rise to the regular arrangement of SF chains by forming hydrogen bonds between SF molecules and consequently the formation of -sheet conformation (as demonstrated
Fig. 6. Schematic diagram of the conformational transition mechanism of SF during the coaxial electrospinning of SF and SS aqueous solutions. The typical amino acid sequence of repetitive motif for the SF is GAGAGS domain [27], while for SS is a characteristic 38-amino acid peptide rich in Ser (SVSSTGSSSNTDSNSNSAGSSTSGGSSTYGYSSNSRDG) [46]. The amino acids with hydrophilic side chains in SF and SS molecules are depicted as white character on a black background.
Y. Hang et al. / International Journal of Biological Macromolecules 51 (2012) 980–986
985
Fig. 7. Non-reversing (upper curve, i) and reversing (lower curve, ii) heat flow results of the coaxially electrospun SF/SS silk fiber (A), and the total heat flow curves for the coaxial SF/SS and pure SF fibers prepared by electrospinning (B).
in Fig. 6(b and c)). Therefore, it is hypothesized from the above analysis that besides serving as an adhesive to bind two SF monofilaments together in cocoon, SS may also play a role in accelerating the reduction of the water content surrounding SF and inducing the conformational transition of SF to -sheet structure.
3.3. Thermal properties of coaxially electrospun SF/SS fiber The thermal properties of coaxially electrospun SF/SS fiber were further investigated by MDSC analysis. The non-reversing (upper curve, i) and reversing (lower curve, ii) heat flow curves of the fiber ranging from 100 to 350 ◦ C were shown in Fig. 7(A). The wide endothermic peak centered at about 125 ◦ C in the non-reversing heat flow curve was due to the heat absorbed by the evaporated water molecules as well as the heat of transformation of the bound water-silk structure to the dry silk structure [48]. The endothermic shift of the baseline without any sharp transition peaks at around 180 ◦ C in the reversing heat flow curve was assigned to the glass transition of silk fibroin [48–51]. Similar shift corresponding to glass transition was also found in the DSC curves of thermotropic liquid crystalline copolyesters [52,53]. Above the temperature of glass transition (Tg ), the non-reversing heat flow curve exhibited a large exothermic peak centered at about 278 ◦ C, which was attributed to the non-isothermal crystallization from the secondary structure of noncrystalline random coil or ␣-helix to -sheet crystal. After the appearance of the crystallization peak, the coaxially electrospun SF/SS fiber exhibited a large endothermic peak at around 290 ◦ C, indicating that the resultant fiber started to degrade [54,55]. Furthermore, Fig. 7(B) showed the total heat flow curves for pure SF fiber and the coaxial SF/SS fiber prepared by electrospinning. It was worth noting that the coaxial SF/SS fiber had a higher degradation temperature (290 ◦ C) than the pure SF fiber (282 ◦ C), which coincided with the results from the TG analysis. According to the method discussed by Jimenez and Li et al. [32–34], some thermal degradation data and kinetic parameters of degradation calculated from the TG results for these two fibers were obtained and shown in Table 2. It was found that, for the coaxial SF/SS fiber, not only its degradation temperature at the maximum weight-loss rate (Tdm ) and residue mass at 500 ◦ C, but also the activation energy (E), the frequency factor (ln Z) and the order (n) of its thermal degradation reaction were all higher than those for the pure SF fiber. On the other hand, the former had a lower maximum weight-loss rate than the latter. These results indicated that the coaxial SF/SS fiber was more thermally stable than the pure SF fiber, which can be attributed to the more stable structure with more -sheet conformation in the coaxial SF/SS fiber.
Table 2 Thermal degradation data and kinetic parameters of degradation calculated from the TG results of regenerated silk fiber mats by electrospinning. Parameters
Fiber sample
Tdm (◦ C) Maximum weight-loss rate (%/min) Residue mass at 500 ◦ C (%) E (kJ/mol) ln Z (1/min) n
Coaxial SF/SS
SF
293 1.86 40.01 45.69 11.66 3.85
286 1.91 38.52 36.87 9.67 3.09
In addition, Fig. 7(B) also displayed the temperature difference in water evaporation of the above two fibers and revealed that the coaxial SF/SS fiber had a lower water evaporation temperature (125 ◦ C) than SF fiber (146 ◦ C). The difference in water evaporation temperature between these two fibers may be concerned with the bound water molecules in the fibers which were relatively hard to be evaporated [48]. It was thought that, for the coaxial SF/SS fiber, the dehydration effect taken by SS considerably decreased the content of the bound water molecules surrounding SF and enabled the transformation of some bound water molecules to free water. This resulted in an easier water loss process and lower temperature for water evaporation in coaxially electrospun SF/SS fiber. 3.4. Mechanical properties of coaxially electrospun SF/SS fiber The mechanical properties of the coaxial SF/SS and pure SF fiber mats prepared by electrospinning were shown in Fig. 8 and Table 3. It was found that the mechanical properties of the electrospun fiber mats were affected obviously by the covering of SS. Compared with the SF fiber mat, the average breaking strength and breaking energy of coaxial SF/SS fiber mat were increased by nearly 82% (to 1.93 MPa) and 93% (to 7.21 J/kg), respectively. One of the main reasons for the improvement of the mechanical properties of the SF/SS fiber mat is the contribution of SS in inducing more -sheet conformation of the SF molecules in the coaxially electrospun SF/SS fiber.
Table 3 Mechanical properties of regenerated silk fiber mats by electrospinning. Fiber sample
Breaking elongation (%)
Breaking strength (MPa)
Breaking energy (J/kg)
a. Coaxial SF/SS b. SF
1.20 ± 0.55 0.96 ± 0.16
1.93 ± 0.40 1.06 ± 0.23
7.21 ± 4.14 3.74 ± 1.75
986
Y. Hang et al. / International Journal of Biological Macromolecules 51 (2012) 980–986
Fig. 8. Stress–strain curves of the regenerated silk fiber mats by electrospinning: (a) coaxial SF/SS and (b) SF.
4. Conclusions This work has demonstrated the feasibility of utilizing the coaxial electrospinning technique to prepare regenerated silk fibroin/silk sericin (SF/SS) fibers in order to mimic the compositions and core–shell structure of the silkworm silk. Based on the morphological analysis of the coaxially electrospun SF/SS fibers, we have acquired useful understanding about the effects of several processing parameters on the formation of core–shell fibers during the coaxial electrospinning. Investigation into the effect of the core dope (SF solution) flow rate (Qc ) revealed that the average fiber diameter of the composite fibers exhibited a Qc -independent trend. However, with a quadrupled increase in Qc , the inner core fiber diameter had an obvious increase and the diameter nonuniformity and eccentricity of the core fiber became more serious. In this work, the Qc of 6 L/min was suitable for the fabrication of coaxial SF/SS fibers with core–shell structure and better electrospinnability, and the corresponding optimum applied voltage (V) was about 40 kV. Moreover, compared with the electrospun pure SF fibers, the coaxial SF/SS fibers had more stable structure with more -sheet (and related intermediate) conformation and accordingly better thermostability and mechanical properties, since SS had a higher content of hydrophilic amino acid side chains and played significant roles in dehydrating SF molecules and inducing the conformational transition of SF to -sheet structure. Acknowledgments This work is supported by the National Natural Science Foundation of China (81170641), Specialized Research Fund for the Doctoral Program of Higher Education (200802550001), Innovation Program of Shanghai Municipal Education Commission (12ZZ065), Shanghai Rising-Star Program (12QA1400100) and the Fundamental Research Funds for the Central Universities. References [1] Y. Shen, M.A. Johnson, D.C. Martin, Macromolecules 31 (1998) 8857–8864. [2] A. Matsumoto, J. Chen, A.L. Collette, U.J. Kim, G.H. Altman, P. Cebe, D.L. Kaplan, Journal of Physical Chemistry B 110 (2006) 21630–21638. [3] H.J. Jin, J. Park, V. Karageorgiou, U.J. Kim, R. Valluzzi, D.L. Kaplan, Advanced Functional Materials 15 (2005) 1241–1247. [4] U.J. Kim, J. Park, H.J. Kim, M. Wada, D.L. Kaplan, Biomaterials 26 (2005) 2775–2785. [5] U.J. Kim, J.Y. Park, C.M. Li, H.J. Jin, R. Valluzzi, D.L. Kaplan, Biomacromolecules 5 (2004) 786–792. [6] J.P. Yan, G.Q. Zhou, D.P. Knight, Z.Z. Shao, X. Chen, Biomacromolecules 11 (2010) 1–5.
[7] J.X. He, Y.R. Qin, S.Z. Cui, Y.Y. Gao, S.Y. Wang, Journal of Materials Science 46 (2011) 2938–2946. [8] K.A. Trabbic, P. Yager, Macromolecules 31 (1998) 462–471. [9] K.H. Lee, D.H. Baek, C.S. Ki, Y.H. Park, International Journal of Biological Macromolecules 41 (2007) 168–172. [10] E. Marsano, P. Corsini, C. Arosio, A. Boschi, M. Mormino, G. Freddi, International Journal of Biological Macromolecules 37 (2005) 179–188. [11] S. Zarkoob, D.H. Reneker, R.K. Eby, S.D. Hudson, D. Ertley, W.W. Adams, Polymer Preprints 39 (1998) 244–245. [12] J. Ayutsede, M. Gandhi, S. Sukigara, M. Micklus, H.E. Chen, F. Ko, Polymer 46 (2005) 1625–1634. [13] Y. Kawahara, A. Nakayama, N. Matsumura, T. Yoshioka, M. Tsuji, Journal of Applied Polymer Science 107 (2008) 3681–3684. [14] F. Zhang, B. Zuo, Z. Fan, Z. Xie, Q. Lu, X. Zhang, D.L. Kaplan, Biomacromolecules 13 (2012) 798–804. [15] H.J. Jin, S.V. Fridrikh, G.C. Rutledge, D.L. Kaplan, Biomacromolecules 3 (2002) 1233–1239. [16] W.H. Park, L. Jeong, D.I. Yoo, S. Hudson, Polymer 45 (2004) 7151–7157. [17] S. Putthanarat, R.K. Eby, W. Kataphinan, S. Jones, R. Naik, D.H. Reneker, B.L. Farmer, Polymer 47 (2006) 5630–5632. [18] H. Wang, Y.P. Zhang, H.L. Shao, X.C. Hu, Journal of Materials Science 40 (2005) 5359–5363. [19] J.X. Zhu, Y.P. Zhang, H.L. Shao, X.C. Hu, Polymer 49 (2008) 2880–2885. [20] H. Cao, X. Chen, L. Huang, Z.Z. Shao, Materials Science and Engineering C 29 (2009) 2270–2274. [21] G.H. Altman, F. Diaz, C. Jakuba, T. Calabro, R.L. Horan, J. Chen, H. Lu, J. Richmond, D.L. Kaplan, Biomaterials 24 (2003) 401–416. [22] N. Kato, S. Sato, A. Yamanaka, H. Yamada, N. Fuwa, M. Nomura, Bioscience, Biotechnology, and Biochemistry 62 (1998) 145–147. [23] X. Zhang, M.M.R. Khan, T. Yamamoto, M. Tsukada, H. Morikawa, International Journal of Biological Macromolecules 50 (2012) 337–347. [24] M. Sasaki, H. Yamada, N. Kato, Nutrition Research 20 (2000) 1505–1511. [25] S. Zhaorigetu, M. Sasaki, H. Watanabe, N. Kato, Bioscience, Biotechnology and Biochemistry 65 (2001) 2181–2186. [26] M.L. Gimenes, L. Liu, X. Feng, Journal of Membrane Science 295 (2007) 71–79. [27] C.J. Fu, Z.Z. Shao, V. Fritz, Chemical Communications 43 (2009) 6515–6529. [28] W. Wei, Y.P. Zhang, H.L. Shao, X.C. Hu, Journal of AOAC International 93 (2010) 1143–1147. [29] Y.C. Hang, Y.P. Zhang, Y. Jin, H.L. Shao, X.C. Hu, Journal of Materials Research 26 (2011) 2931–2937. [30] L. Zhou, X. Chen, Z.Z. Shao, Y.F. Huang, D.P. Knight, Journal of Physical Chemistry B 109 (2005) 16937–16945. [31] P. Zhou, X. Xie, D.P. Knight, X.H. Zong, F. Deng, W.H. Yao, Biochemistry 43 (2004) 11302–11311. [32] A. Jimenez, V. Berenguer, J. Lopez, A. Sanchez, Journal of Applied Polymer Science 50 (1993) 1565–1573. [33] X.G. Li, M.R. Huang, Polymer Degradation and Stability 64 (1999) 81–90. [34] X.G. Li, M.R. Huang, Y.Q. Lu, M.F. Zhu, Journal of Materials Chemistry 15 (2005) 1343–1352. [35] M. Wang, J.H. Yu, D.L. Kaplan, G.C. Rutledge, Macromolecules 39 (2006) 1102–1107. [36] C. Wang, K.W. Yan, Y.D. Lin, P.C.H. Hsieh, Macromolecules 43 (2010) 6389–6397. [37] C. Wang, Y.W. Cheng, C.H. Hsu, H.S. Chien, S.Y. Tsou, Journal of Polymer Research 18 (2011) 111–123. [38] Z.C. Sun, E. Zussman, A.L. Yarin, J.H. Wendorff, A. Greiner, Advanced Materials 15 (2003) 1929–1932. [39] J. Perez-Rigueiro, M. Elices, J. Llorca, C. Viney, Journal of Applied Polymer Science 82 (2001) 1928–1935. [40] A.K. Moghe, B.S. Gupta, Polymer Reviews 48 (2008) 353–377. [41] J.M. Deitzel, J. Kleinmeyer, D. Harris, N.C. Beck Tan, Polymer 42 (2001) 261–272. [42] P. Monti, G. Freddi, A. Bertoluzza, N. Kasai, M. Tsukada, Journal of Raman Spectroscopy 29 (1998) 297–304. [43] P. Monti, P. Taddei, G. Freddi, T. Asakura, M. Tsukada, Journal of Raman Spectroscopy 32 (2001) 103–107. [44] X.G. Li, L.Y. Wu, M.R. Huang, H.L. Shao, X.C. Hu, Biopolymers 89 (2008) 497–505. [45] R. Tuma, P.E. Prevelige, G.J. Thomas, Biochemistry 35 (1996) 4619–4627. [46] H. Teramoto, A. Kakazu, K. Yamauchi, T. Asakura, Macromolecules 40 (2007) 1562–1569. [47] H. Teramoto, T. Kameda, Y. Tamada, Bioscience, Biotechnology, and Biochemistry 72 (2008) 3189–3196. [48] X. Hu, D. Kaplan, P. Cebe, Macromolecules 41 (2008) 3939–3948. [49] X. Hu, D. Kaplan, P. Cebe, Thermochimica Acta 461 (2007) 137–144. [50] A. Motta, L. Fambri, C. Migliaresi, Macromolecular Chemistry and Physics 203 (2002) 1658–1665. [51] N. Agarwal, D.A. Hoagland, R.J. Farris, Journal of Applied Polymer Science 63 (1997) 401–410. [52] X.G. Li, Z.L. Zhou, X.G. Wu, T. Sun, Journal of Applied Polymer Science 51 (1994) 1913–1921. [53] X.G. Li, M.R. Huang, G.H. Guan, T. Sun, Journal of Applied Polymer Science 66 (1997) 2129–2138. [54] G. Freddi, G. Pessina, M. Tsukada, International Journal of Biological Macromolecules 24 (1999) 251–263. [55] Y. Tao, W. Xu, Y. Yan, Y. Cao, Polymer International 61 (2012) 760–767.