Wet-Spinning of Osmotically Stressed Silk Fibroin - American

2086

Biomacromolecules 2009, 10, 2086–2091

Wet-Spinning of Osmotically Stressed Silk Fibroin Sungkyun Sohn* and Samuel P. Gido* Department of Polymer Science and Engineering, University of Massachusetts, Amherst, Massachusetts 01003 Received February 8, 2009; Revised Manuscript Received May 29, 2009

Based on the phase diagram constructed for water-silk fibroin-LiBr using the osmotic stress method, wet-spinning of osmotically stressed, regenerated Bombyx mori silk fibroin was performed, without the necessity of using expensive or toxic organic solvents. The osmotic stress was applied to prestructure the regenerated silk fibroin molecule from its original random coil state to a more oriented state, manipulating the phase of the silk solution in the phase diagram before the start of spinning. Various starting points for spinning were selected from the phase diagram to evaluate the spinning performance and also physical properties of fibers produced. Monofilament fiber with a diameter of 20 µm was produced. It was found that the fibers whose starting point in the phase diagram were around the phase boundary between silk I and silk II, at very low LiBr concentrations, showed the best spinning process stability and physical properties. This underpins the prediction that the enhanced control over structure and phase behavior using the osmotic stress method helps improve the physical properties of wetspun regenerated silk fibroin fibers.

Introduction The exceptional mechanical properties of silk fibers and the efficiency of silk spinning in nature have produced a great deal of effort to understand the nature of the silk process as it proceeds from the biosynthesis of silk fibroin to the spinning of the fiber. However, many aspects of the process still elude investigators. Natural silk fibers of domestic Bombyx mori silkworm are semicrystalline, with a high degree of molecular orientation.1 These natural fibers are formed through an ambient pressure and temperature process,2 and although natural silk is spun from an aqueous solution in vivo, the resulting fiber is water insoluble due to an irreversible phase transition during the spinning process. An oriented intermediate state, possibly liquid crystalline, of silk in aqueous solution, with a high degree of precrystalline molecular orientation, has long been postulated and could explain the ability of this natural process to transform a molecule that exists as a random coil in aqueous solution into an insoluble, highly oriented fiber.1-6 Artificial production of such fibers would be extremely desirable because variations of the physical properties, diameter, shape of cross-section, as well as morphology and structure of fibers are something that cannot be achieved by the spinning process in vivo with the currently available technology. Therefore, it is essential to understand the processing conditions that induce the hierarchical self-assembly of silk fibroin molecules that leads to the excellent physical properties. So far, many researchers have focused on the regeneration of silk fibers, including spider silk, using expensive or toxic organic solvents such as N-methylmorpholine N-oxide (NMMO),7 trifluoro-acetic acid,8 hexafluoroisopropanol (HFIP), or hexafluoroacetone (HFA).9-15 We note that force-reeling directly from silkworms16 or using a warm aqueous ammonium sulfate coagulation bath17 yields fibers with exceptional mechanical properties. However, the most desirable approach would be exactly simulating the natural process, where water is the only solvent. In our previous study, it was shown that the osmotic stress method can provide a means for the direct * To whom correspondence should be addressed. E-mail: sksohn@ mail.pse.umass.edu (S.S.); [email protected] (S.P.G).

investigation of the microscopic and thermodynamic details of intermolecular interactions of silk fibroin molecules in aqueous salt solution5,18,19 and that the osmotic pressure can effectively control the silk fibroin composition in the aqueous LiBr solution, as shown in Figure 1. Also, a partial ternary phase diagram of water-silk fibroin-LiBr was successfully constructed based on X-ray results, and it was suggested that designing a new wetspinning route, where the prespun silk fibroin molecules were prestructured to form a hydrated intermediate structure by applying osmotic stress, would be desirable.5 In this study, a process for the artificial wet-spinning of Bombyx mori silk fibroin in osmotically stressed aqueous salt solution was confirmed. Optical microscopy, confocal Raman scattering, and X-ray diffraction are used for the characterization of the regenerated silk fibers prepared.

Experimental Section Materials. Silk cocoons of Bombyx mori silkworms were received from the National Institute of Agrobiological Sciences, Tsukuba, Japan. Sodium carbonate (Aldrich), Marseille soap (French soaps), lithium bromide (Aldrich), poly(ethylene glycol) (PEG; Mw ) 8000 g/mol, Fluka), and methyl alcohol (Pharmco) were used as received. Degumming and Silk Solution Preparation. Degumming and silk solution preparation were done as described in ref 5. Pure silk fibroin was obtained by removing sericin, a gummy binding protein coating the silk fibroin filaments, through the degumming process. The silk cocoons of Bombyx mori were boiled in four changes of water. First, the cocoons were treated with pure boiling water for 30 min and then rinsed thoroughly in running deionized water. Second, the cocoons were boiled in aqueous 1 × 10-4 wt % Na2CO3 solution and 7 × 10-5 wt % Marseille soap for 2 h. Third, the cocoons were boiled again in aqueous 7 × 10-5 wt % Na2CO3 solution in the presence of the 7 × 10-5 wt % soap for 1 h. Fourth, the cocoons were washed for a final hour in pure boiling water. Then the degummed fibroin was dried at 50 °C in a vacuum oven for 1 day. The solid (silk)-to-liquid ratio was kept as 1:150 throughout the degumming process. Dried, degummed silk fibroin was then dissolved in an aqueous 9.0 M LiBr solution to yield an 18 wt % silk solution. For comparison, natural Bombyx mori silk monofilament fiber was separately prepared by the same degumming procedure described above.

10.1021/bm900169z CCC: $40.75  2009 American Chemical Society Published on Web 07/02/2009

Wet-Spinning of Osmotically Stressed Silk Fibroin

Biomacromolecules, Vol. 10, No. 8, 2009

2087

Figure 1. Composition control of silk fibroin by the osmotic stress method. Table 1. Initial and Final Compositions of Osmotically Stressed Silk Fibroin Solutions and the Silk Structure Corresponding to Each Composition stressing solution

silk subphase composition

designation

LiBr [M]

PEG (wt %)

Log10 [Π(dyn/cm2)]a

LiBr (wt %)

silk (wt %)

water (wt %)

structureb

9M-10 7M-20 7M-30 5M-20 5M-25 5M-30 3M-15 3M-20 3M-25 3M-30 1M-15 1M-20 1M-25 1M-30

9.0 7.0 7.0 5.0 5.0 5.0 3.0 3.0 3.0 3.0 1.0 1.0 1.0 1.0

10 20 30 20 25 30 15 20 25 30 15 20 25 30

6.25 6.89 7.34 6.89 7.13 7.34 6.61 6.89 7.13 7.34 6.61 6.89 7.13 7.34

44.9 34.9 35.2 24.8 25.0 25.1 15.0 14.9 15.2 15.1 5.6 5.2 5.3 5.4

5.1 15.1 22.0 20.0 25.1 30.0 18.0 24.0 31.1 37.1 20.9 29.0 36.9 42.9

50.0 50.0 42.8 55.2 49.9 45.1 67.0 61.1 54.1 47.8 73.5 65.8 57.8 51.7

R R + S-I S-I S-I S-I S-I + S-II S-I S-I S-I + S-II S-II S-I S-I + S-II S-II S-II

a Ref 18, Π (dyn/cm2) ) -1.31 × 106G2T + 141.8 × 106G2 + 4.05 × 106G. G ) PEG wt %/(100 - PEG wt %), T ) 25. S-II: silk II.

Preparation of Osmotically Stressed Silk Solution. Table 1 lists the full matrix of silk solutions prepared for spinning, including the compositions of both the stressing subphase and the silk containing subphase.5 For the stressing subphase, samples were selectively prepared among combinations of PEG concentrations, 10, 15, 20, 25, and 30 wt %, with LiBr concentrations of 1.0, 3.0, 5.0, 7.0, and 9.0 M. This provides 14 stressing subphases of varying osmotic stress and chaotrope (salt) concentration. Based upon our previously determined phase diagram,5 10 compositions were targeted in the intermediate silk I phase or its boundaries. In 50 mL tubes, each stressing solution is placed in proximity to a silk containing aqueous solution. Initially, the silk solution placed in each tube was the same, an 18 wt % silk stock solution in 9.0 M LiBr. A 10 mL aliquot of this stock solution was added to each Eppendorf tube, followed by the addition of about 40 mL of the corresponding PEG/LiBr stressing solution. Because of initial differences in salt concentration and osmotic pressure between the stock solution and the various stressing solutions, water and salt ions are exchanged across the subphase boundary. This changes the concentrations of the various components in both the silk subphase and the stressing solution subphase. To equilibrate each silk solution (located in the bottom of the 50 mL tube) to the desired constant osmotic

b

R: random coil; S-I: silk I;

pressure, the PEG/LiBr stressing solution (located in the upper part of the 50 mL tube) was exchanged twice per day for two weeks with fresh solution of the desired composition. Equilibration of the subphase must be performed with extra care so that the silk solution and the stressing solution, in the 50 mL tube, are not mixed, especially when the stressing solution is exchanged. In this way, the stressing subphase was held at approximately constant osmotic pressure and salt concentration, while the silk subphase was driven to equilibrate with this set of stressing solution conditions. After the two week period of equilibration, the separation of the silk subphase in the bottom of the 50 mL tube became more clearly visible. The compositions of the silk solution subphases once equilibrium has been established are also listed in Table 1 adjacent to the compositions of the corresponding stressing subphase. Calculation of Silk Subphase Composition. After equilibration against stressing solutions, the compositions (water, silk, and salt) of the various silk containing subphases were calculated as follows: water content was calculated based on the weight difference before and after drying each silk subphase sample. Silk content relative to salt was calculated by dialyzing a measured amount of sample against deionized water using a dialysis membrane (SpectraPor Cellulose Ester, Mw cutoff

2088


Figure 2. Schematic illustration of the wet-spinning process for the preparation of osmotically stressed regenerated silk fibers.

1000, Spektrum), followed by drying out the water and weighing the silk fibroin remaining. Wet Spinning. Figure 2 shows a schematic illustration of the wetspinning apparatus for osmotically stressed, regenerated silk fibers. After equilibration, osmotically stressed silk fibroin solution was filtered by size 4 filter paper to remove impurities such as small particles of precipitated fibroin. Then the solution was transferred to a spin pack equipped with a stainless steel spinneret (Nippon Nozzle) for circular monofilament with a diameter of 0.25 mm and a capillary length of 1.5 mm. Extrusion was driven by high-pressure nitrogen. The spinning apparatus was installed such that as-spun fiber leaving the nozzle passed through a 1.0 cm air gap (controlled to be 25 °C) before diving into the room temperature (25 °C) coagulation bath of methanol and water (1:1 by volume). X-ray Diffraction. Wide-angle X-ray diffraction (WAXD) was performed using an instrument from Molecular Metrology Inc., equipped with a focusing multilayer monochromator (Osmic MaxFlux) with a beam wavelength of 0.154 nm. The beam was collimated with three pinholes. To record wide angle profiles, a 20 × 20 cm2 Fuji BAS 2500 ST-VA image plate with a hole in the center was inserted into the beam path. Laser Confocal Raman Spectroscopy. Laser confocal Raman spectroscopy (LabRam HR800 Raman microscope, Jobin Yvon Co.) was used to characterize structural development of the regenerated silk fiber. This instrument can achieve a spatial resolution of ∼1 µm2 area. The spectral resolution was kept at 1 cm-1. Optical Microscopy. Polarizing optical microscopy was performed with a Zeiss Axiovert S100TV inverted polarizing microscope equipped with an LC Pol-Scope retardance imaging system (CRI, Boston, MA). Mechanical Test. Mechanical properties of the natural and regenerated silk fibers were all measured using a tensile testing machine (Instron 5564) at a temperature of 20 °C and a relative humidity of 65%. All samples were equilibrated in a controlled temperature and humidity environment for at least 24 h before testing. Each sample length was 5 cm and the cross-head speed was 4 mm/min. Five measurements were performed for each sample.

Results and Discussion Wet-spinning silk fibers from osmotically stressed, regenerated Bombyx mori silk fibroin was planned foreseeing several technical advantages compared with the fibers produced without applying osmotic stress. In the natural silk spinning process, a silk I like structure is thought to be an intermediate between random coil and silk II.5,20-22 As indicated by the phase diagram reported in our previous study,5 if spinning starts from the

Sohn and Gido

random coil state, conversion to silk II by means of changing water and salt concentration always involves passing through two irreversible phase transitions (random coil to silk I, and silk I to silk II), making it a long route, which can possibly contribute to unevenness of the resulting fiber. As nature utilizes the silk I like intermediate to prealign molecules and to shed water, one can envision using osmotic stress to prestructure silk fibroin molecules in an artificial spinning process that mimics nature’s use of silk I, shortening the pathway from the start to the end of spinning process. The schematic illustration of the suggested new spinning route as well as the wet-spinning apparatus is given in Figure 2. For each run, about 5 mL of each silk solution prepared under osmotic stress was transferred to the spinpack. PEG 8K was thoroughly excluded from flowing into the spinning stream by removal of the boundary area between the two subphases as well as the stressing subphase itself. As the silk solution gets extruded through the nozzle and solidifies in the coagulation bath filled with methanol and water (1:1 by volume), the tip of the solidified fiber is held by tweezers and put onto the correct fiber path. Overall, 14 starting points were selected from the phase diagram, which are broadly distributed in the silk I region of the phase diagram at relatively low LiBr concentrations. The intention of this study was to find the optimum starting point for fiber spinning. These points are listed with composition and phase data in Table 1. The four points in the random coil and silk II phases (9M-10, 3M-30, 1M-25, and 1M-30) were added for comparison. Spinning speed was determined by the first Godet roller installed outside the coagulation bath. Spinning speed was subject to the process stability (which was frequently hindered by fiber breakage due to insufficient solidification or mechanical conditions, etc.), and up to 50 cm/min, which is the spinning speed of the Bombyx mori silkworm, was achieved in the case of 1M-20. For silk fibroin solutions that contained more than 3 M LiBr (9M-10, 7M-20, 7M-30, 5M-20, 5M-25, and 5M-30) or those started from the silk II phase (3M-25, 3M-30, 1M-25, and 1M-30), very brittle “dry-noodle-like” fibers were obtained and only with extreme care at very mild spinning conditions (spinning speed slower than 20 cm/min). The choice of methanol and water bath over traditional methanol only bath was made to improve process stability. The methanol only bath made the rate of structural transition exceed the acceptable limit for spinning and drawing, and the desired water only bath hardly provide any processing window. It is considered that adjusting pH and/or addition of appropriate salt(s) may potentially enable the ideal aqueous bath suitable for wet-spinning of osmotically stressed silk fibroin with no or reduced use of methanol.17,23 In general, the rate of structural transition in the coagulation bath must be fast enough to ensure fiber formation, but it also has to be slow enough to avoid the formation of brittle fibers.17 Ideally, it is desirable to expel all lithium and bromide ions out of the fiber mass, as soon as the fiber dips into the coagulation bath, allowing fast crystallization of silk fibroin molecules. For the samples with high LiBr concentrations, the limits of this ion removal process is thought to be the main reason for instability during the spinning process, frequent yarn breakage, and resulting fibers with very poor physical properties. Four samples were started from the silk I phase or its boundary with silk II (3M-15, 3M-20, 1M-15, and 1M-20). In these cases the spinning process was relatively stable and fibers with better physical properties were obtained, although the properties were still not comparable to those of natural silk fibroin fiber. Fibers spun from 1M-20 showed the best process stability and physical properties. It has been postulated that there exists a hydrated


Figure 3. Evaluation of spinnability from the selected starting points for the spinning of osmotically stressed regenerated silk fibroin: (a) area where the spinning process was relatively stable and (b) area where the spinning process was very unstable.


2089

Figure 5. Stress-strain curves of (a) natural Bombyx mori silk fibroin fiber, (b) regenerated silk fibroin fiber spun from the osmotically stressed silk solution (1M-20) spun at 50 cm/min and 1.10 draw ratio, and (c) regenerated silk fibroin fiber spun from 9 M LiBr silk solution (9M-10) spun at 20 cm/min and 1.00 draw ratio. Table 2. Young’s Modulus and Strain at Break of Natural and Regenerated Bombyx mori Silk Fibroin Fibers regenerated

Young’s modulus (cN/dtex) strain at break (%)

Figure 4. Regenerated silk fibroin fiber spun from the osmotically stressed silk solution (1M-20): (a) before drawing and (b) after takeup.

intermediate structure, or silk I-like structure, of silk fibroin.5,20-22,24,25 Our spinning results, where silk fibers made from an osmotically stressed intermediate silk I-state showed better process stability and physical properties over those started from either the random coil state or the crystallized state, supports this postulate. Figure 3 groups the spinning starting points into two groups: (a) those giving a stable spinning process and (b) those giving unstable spinning. As-spun fibers in this study were 20-50 µm in diameter, depending on spinning conditions. Preparation of fibers with a diameter smaller than 20 µm was severely hindered by frequent yarn breakage encountered during the spinning process. The diameters of asspun fibers were measured by optical microscopy, and birefringence was observed. Figure 4a,b shows the point of elongation of a fiber due to mechanical drawing, and a drawn fiber, respectively. These are polarizing optical micrographs of spun fibers from the silk solution 1M-20. Although a natural drawing actually occurs during the solidification process from the tip of the nozzle to the first guide roller, sometimes even at the locations close to the first Godet roller, “drawing” in this study

natural

from 1M-20

from 9M-10

86.6 19.2

44.5 7.6

18.1 2.2

implies mechanical drawing with a draw ratio defined by the speed difference between the first and the second Godet rollers. Figure 5 shows a comparison between the stress-strain curves of natural Bombyx mori silk fiber, the regenerated silk fiber spun from the silk solution 9M-10, and the regenerated silk fiber spun from the silk fibroin solution 1M-20. Young’s modulus for the natural Bombyx mori silk fiber, the regenerated silk fiber spun from the silk solution 9M-10, and the regenerated silk fiber spun from the silk solution 1M-20 were calculated from Figure 5, and these values are shown in Table 2. It is clear that the fiber made from 9M-10 does not possess any physical properties that could be desirable for textile fibers, not to mention the difficulty of overcoming frequent yarn breakages and process instability when it was spun. Figure 6 is a plot of the maximum breaking stress versus strain at break. Spinning conditions were adjusted to produce these fibers with different elongations. It is also clear from the figure that fibers spun from 1M-20 not only have more respectable physical properties, but also come with a much wider processing window than those from 9M-10; that is, 9M-10 could

Figure 6. Maximum breaking stress vs strain at break of regenerated silk fibroin fiber spun from osmotically stressed silk solution (1M-20).

2090


Figure 7. Breaking stress vs fiber draw ratio of (a) regenerated silk fibroin fiber spun from osmotically stressed silk solution (1M-20) and (b) regenerated silk fibroin fiber spun from 9 M LiBr silk solution (9M-10).

Sohn and Gido

Figure 10. Raman spectra of (a) 1M-20 before spinning, (b) regenerated silk fibroin fiber spun from 1M-20, and (c) natural Bombyx mori silk fibroin fiber.

Figure 8. Strain at break vs fiber draw ratio of (a) regenerated silk fibroin fiber spun from osmotically stressed silk solution (1M-20) and (b) regenerated silk fibroin fiber spun from 9 M LiBr silk solution (9M-10).

Figure 9. Silk factor of regenerated silk fibroin fiber spun from osmotically stressed silk solution (1M-20).

only be spun to have a very limited range of physical properties (∆stress ) 0.08 cN/dtex, ∆strain ) 2.20%), while spinning conditions could be more flexibly adjusted in case of 1M-20, giving a wider range of physical properties (∆stress ) 0.54 cN/ dtex, ∆strain ) 9.24%). Figure 7g stress versus fiber draw ratio of the fibers spun from 1M-20 and 9M-10, while a plot of strain at break versus fiber draw ratio of the fibers spun from 1M-20 and 9M-10 is shown in Figure 8. The fibers from 1M-20 could be spun to have wider range of physical properties than the fibers spun from 9M-10, and 9M-10 was limited to lower spinning speed. While it is known that the tensile properties of brittle regenerated fibers may be modified by a wet-stretching

Figure 11. Wide angle X-ray diffraction patterns of a regenerated silk fibroin fiber spun from 1M-20 and a natural Bombyx mori silk fibroin fiber: (a) 1M-20 and (b) natural silk.

process,26 immersion of Godet rollers into the coagulation bath did not improve physical properties of the fiber spun from 9M10, and the draw ratio could not be increased. Comparison of the silk factor, defined as the breaking strength of a fiber in gram-force per denier (denier is defined as the number of grams for a fiber of 9000 m length.) multiplied by square root of elongation in percent,27 is shown in Figure 9. Although not as high as the calculated silk factor for natural Bombyx mori silk fiber, which is over 20, fibers spun from 1M-20 with values of about 3 show a large improvement over those spun from 9M10. 1M-20 shows useful properties, considering that the breaking stress of regenerated Bombyx mori silk fibroin fibers of similar diameter prepared using fluorine-based solvents were only about


double that of the fiber spun from 1M-20.12,13 The gap in properties between osmotically stressed regenerated silk fibroin fibers and natural silk fibers could possibly be overcome by improving molecular weight and molecular weight distribution in the regenerated silk fibroin solution, and adjusting spinning, drawing, and postspinning treatment conditions in the future. Figure 10 shows Raman spectra of 1M-20 before and after spinning and of natural Bombyx mori silk fibroin fiber. The amide I band at 1665 cm-1, from CdO s in β-sheets, sharpens as 1M-20 transforms to fiber through crystallization, although not to the degree of that of the natural silk fiber. The amide III band at 1263 cm-1, from N-H in-plane bending and C-N stretching, also shows the same trend.12,28-31 From the smaller width of these bands in fibers of 1M-20 than those of 1M-20 solution, it is evident that β-sheet content in the fibers of 1M20 is substantially higher than the solution. Wide angle X-ray diffraction of a regenerated silk fibroin fiber (1M-20) and a natural silk fiber are shown in Figure 11. A similarity between the Raman spectra and WAXD patterns of fiber spun from 1M20 and the natural silk fibroin fiber indicates the structural similarity of these two fibers. Consequently, these results suggest that osmotically stressed, prestructured silk fibroin fibers (especially 1M-20) are well-oriented along the fiber axis by the wet-spinning process.

Conclusion Wet-spinning of osmotically stressed Bombyx mori silk fibroin was performed. Spinning processability and mechanical test results indicate a pathway dependence of the wet-spinning process. By applying osmotic stress, the composition and corresponding crystalline structure of a regenerated silk solution were precisely controlled before spinning, and a new route mimicking nature’s way of using intermediate structures was designed. The effect of prestructuring silk fibroin molecules was critical on improving physical properties of the fiber as well as wet-spinning processability. The composition at around the boundary between silk I and silk II was found to be the most promising starting point for spinning. Fibers up to a 20 µm diameter were produced by this method. Acknowledgment. NSF MRSEC is acknowledged for financial support for this research. The authors thank Dr. T. Hata at the National Institute of Agrobiological Sciences, Japan, for supplying raw materials. Prof. E. D. T. Atkins (Univ. of Bristol, U.K.), and Dr. C. B. Stanley (Oak Ridge National Laboratory) are acknowledged for helpful discussions.

References and Notes (1) Magoshi, J.; Magoshi, Y.; Nakamura, S. J. Appl. Polym. Sci.: Appl. Polym. Symp. 1985, 41, 187.


2091

(2) Viney, C.; Bell, F. I. Curr. Opin. Solid State Mater. Sci. 2004, 8, 165. (3) Kerkam, K.; Viney, C.; Kaplan, D. L.; Lombardi, S. Nature 1991, 349, 596. (4) Magoshi, J. ; Magoshi, Y.; Nakamura, S. In Silk Polymers: Material Science and Biotechnology; Kaplan, D. L., Adams, W. W., Farmer, B., Viney, C. , Eds.; American Chemical Society: Washington, DC, 1994. (5) Sohn, S.; Strey, H. H.; Gido, S. P. Biomacromolecules 2004, 5, 751. (6) Asakura, T.; Umemura, K.; Nakazawa, Y.; Hirose, H.; Higham, J.; Knight, D. P. Biomacromolecules 2007, 8, 175. (7) Marsano, E.; Corsini, P.; Arosio, C.; Boschi, A.; Mormino, M.; Freddi, G. Int. J. Biol. Macromol. 2005, 37, 179. (8) Ha, S. W.; Tonelli, A. E.; Hudson, A. E. Biomacromolecules 2005, 6, 1722. (9) Lock, R. L. U.S. Patent 5,252,285, 1993. (10) Matsumoto, K.; Uejima, H.; Iwasaki, T.; Sano, Y.; Sumino, H. J. Appl. Polym. Sci. 1996, 60, 507. (11) Mathur, A. B.; Tonelli, A.; Rathke, T.; Hudson, S. Biopolymers 1997, 42, 61. (12) Trabbic, K. A.; Yager, P. Macromolecules 1998, 31, 462. (13) Seidel, A.; Liivak, O.; Jelinski, L. W. Macromolecules 1998, 31, 6733. (14) Liivak, O.; Blye, A.; Shah, N.; Jelinski, L. W. Macromolecules 1998, 31, 2947. (15) Yao, J.; Masuda, H.; Zhao, C.; Asakura, T. Macromolecules 2002, 35, 6. (16) Shao, Z. Z.; Vollrath, F. Nature 2002, 418, 741. (17) Zhou, G.; Shao, Z. Z.; Knight, D. P.; Yan, J.; Chen, X. AdV. Mater. 2009, 21, 366. (18) Parsegian, V. A.; Rand, R. P.; Fuller, N. L.; Rau, D. C. In Methods in Enzymology; Packer, L., Ed.; Academic Press: Orlando, FL, 1986; Vol. 127, p 400. (19) Parsegian, V. A.; Rand, R. P.; Rau, D. C. In Methods in Enzymology; Johnson, M. L.;, Ackers, G. K. , Eds.; Academic Press: New York, NY, 1995; Vol. 259, p 43. (20) Jin, H.-J.; Kalpan, D. L. Nature 2003, 424, 1057. (21) Ro¨ssle, M.; Panine, P.; Urban, V. S.; Riekel, C. Biopolymers 2004, 74, 316. (22) Martel, A.; Burghammer, M.; Davies, R. J.; Di Cola, E.; Vendrely, C.; Riekel, C. J. Am. Chem. Soc. 2008, 130, 17070. (23) Xie, F.; Zhang, H.; Shao, H.; Hu, X. Int. J. Biol. Macromol. 2006, 38, 284. (24) Atkins, E. D. T. 3rd International Silk Conference, Montreal, QC, 2003, presentation. (25) Atkins, E. D. T. University of Bristol: Bristol, U.K. Personal communication, 2005. (26) Plaza, G. R.; Corsini, P.; Perez-Rigueiro, J.; Marsano, E.; Guinea, G. V.; Elices, M. J. Appl. Polym. Sci. 2008, 109, 1793. (27) Yoshimoto, M.; Ohwaki, S. U.S. Patent 4,791,026, 1988, p 791. (28) Magoshi, J.; Mizuide, M.; Magoshi, Y.; Takahashi, K.; Kubo, M.; Nakamura, S. J. Polym. Sci., Polym. Phys. Ed. 1979, 17, 515. (29) Monti, P.; Freddi, G.; Bertoluzza, A.; Kasai, N.; Tsukada, M. J. Raman Spectrosc. 1998, 29, 297. (30) Rousseau, M.; Lefevre, T.; Beaulieu, L.; Asakura, T.; Pezolet, M. Biomacromolecules 2004, 5, 2247. (31) Qiu, W.; Teng, W.; Cappello, J.; Wu, X. Biomacromolecules 2009, 10, 602.

BM900169Z

Wet-Spinning of Osmotically Stressed Silk Fibroin - American

Recommend Documents