Porous, Aligned and Biomimetic Fibers of Regenerated Silk Fibroin

Subscriber access provided by University of Winnipeg Library

Article

Porous, Aligned and Biomimetic Fibers of Regenerated Silk Fibroin Produced by Solution Blow Spinning Adrian Magaz, Aled D. Roberts, Sheida Faraji, Tatiana R. L. Nascimento, Eliton S Medeiros, Wenzhao Zhang, Ryan D. Greenhalgh, Andreas Mautner, Xu Li, and Jonny J. Blaker Biomacromolecules, Just Accepted Manuscript • DOI: 10.1021/acs.biomac.8b01233 • Publication Date (Web): 02 Nov 2018 Downloaded from http://pubs.acs.org on November 3, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 39 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biomacromolecules

1

Porous, Aligned and Biomimetic Fibers of Regenerated Silk

2

Fibroin Produced by Solution Blow Spinning

3

Adrián Magaz1,2†, Aled D. Roberts1†, Sheida Faraji1, Tatiana R.L. Nascimento3, Eliton

4

S. Medeiros3, Wenzhao Zhang1, Ryan D. Greenhalgh1⊥, Andreas Mautner4, Xu Li2,5*,

5

Jonny J. Blaker1*

6

1

7

Manchester, United Kingdom

8

2

9

Technology and Research (A*STAR), Singapore

Bio-Active Materials Group, School of Materials, The University of Manchester,

Institute of Materials Research and Engineering (IMRE), Agency for Science,

10

3

11

Universidade Federal da Paraíba, João Pessoa, Brazil

12

4

13

Research, University of Vienna, Vienna, Austria

14

5

Laboratory of Materials and Biosystems, Department of Materials Engineering,

Polymer and Composite Engineering Group, Institute of Materials Chemistry and

Department of Chemistry, National University of Singapore, Singapore

15

Present address: ⊥Department of Physiology, Development and Neuroscience,

16

University of Cambridge, United Kingdom

17

18

†Equal author contribution *Corresponding authors:

Jonny J. Blaker

Li Xu

School of Materials

Institute

of

Materials

Research

and 1

ACS Paragon Plus Environment

Biomacromolecules 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The University of Manchester

Engineering (IMRE)

Manchester, M13 9PL, UK

Singapore, 138634

tel. +44(0)161 3063578

tel. +65 64168933

e-mail: [email protected]

e-mail: [email protected]

Page 2 of 39

19 20

Abstract

21

Solution blow spinning (SBS) has emerged as a rapid and scalable technique for the

22

production of polymeric and ceramic materials into micro/nano- fibers. Here, SBS was

23

employed to produce sub-micron fibers of regenerated silk fibroin (RSF) from Bombyx

24

mori (silkworm) cocoons based on formic acid or aqueous systems. Spinning in the

25

presence of vapor permitted the production of fibers from aqueous solutions, and high

26

alignment could be obtained by modifying the SBS set-up to give a concentrated

27

channeled airflow. The combination of SBS and a thermally induced phase separation

28

technique (TIPS) resulted in the production of macro/micro- porous fibers with 3D

29

interconnected pores. Furthermore, a co-axial SBS system enabled a pH gradient and

30

kosmotropic salts to be applied at the point of fiber formation, mimicking some of the

31

aspects of the natural spinning process, fostering fiber formation by self-assembly of the

32

spinning dope. This scalable and fast production of various types of silk based fibrous

33

scaffolds could be suitable for a myriad of biomedical applications.

34

Keywords: Regenerated silk fibroin, solution blow spinning, fibers, porous fibers,

35

aligned fibers, biomimetic spinning

36 37

1. INTRODUCTION

38

Solution blow spinning (SBS) offers an efficient method to produce micro/nano- fibers 2


Page 3 of 39 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biomacromolecules

1,2

39

at a significantly higher rate (about ×100 times faster

) compared to conventional

40

electrospinning. By using pressurized gas as the driving force,2–5 it overcomes some of

41

the drawbacks of electrospinning including the use of electric fields and dense fibrous

42

networks with low porosities;6–9 in addition, fibers can be deposited in situ on virtually

43

any surface.4 To date, a myriad of polymers,3–5 composites 1,10,11 and ceramics 12,13 have

44

been successfully processed into fibers by SBS. The use of micro/nano- fibers is of

45

particular interest in the field of regenerative medicine due to their high surface area to

46

volume ratio and morphological resemblance to the extracellular matrix of native

47

tissues. Fibers can provide adequate topographical and mechanical cues to direct cell

48

behavior, making them suitable for the fabrication of scaffolds for tissue regeneration.14–

49

16

50

In its native state, Bombyx mori silk consists of fibrous protein filaments (silk fibroin)

51

surrounded by a glue-like layer (sericin). Among the wide range of natural materials,

52

silk fibroin has many attractive chemical, physical, and biological properties which

53

make it convenient for regenerative medicine, tissue engineering and therapeutic

54

delivery applications.17,18 Some of these properties include suitable cell-material

55

interactions, tailored biodegradability and oxygen/water permeability.19–22

56

Under natural spinning conditions, as the silk dope flows from the posterior to the

57

anterior division of the silkworm gland, the dope is sheared, water content decreases

58

from 80 to 77%, pH decreases from 6.9 to 4.8 and chaotropic ions are exchanged for

59

kosmotropic ions,23–29 contributing to fiber formation. Silk fibroin typically requires

60

reprocessing from its native state to tune properties including fiber diameter and surface

61

roughness/morphology to suit the intended application15,30. In particular, regenerated

62

silk fibroin (RSF) from aqueous solutions has previously been electrospun by blending

63

with high molecular weight poly(ethylene oxide) (PEO),21,31,32 used as a temporarily 3



Page 4 of 39

64

water-soluble spinning aid, and fully aqueous processes based on high molecular weight

65

silk fibroin dopes have been successfully produced via electrospinning.33,34 Recently,

66

straining flow spinning (SFS) has been used to produce high performance RSF fibers by

67

controlling the flow of the dope through its interaction with a coagulant bath

68

mimicking the balanced mild aqueous chemical environment and moderate shear forces

69

found in natural spinning glands.35 While robust RSF fibers can be mass-produced by

70

SFS, these range between 8-13 μm in diameter. SBS permits faster production rates (5

71

μL.min−1 in SFS compared with >100 μL.min−1 in SBS) and can even achieve thinner

72

fibers at the nanoscale.

73

The aim of this study is not to produce high performance robust fibers nor retain the

74

hierarchical structures of natural silks, but rather show that SBS is a suitable technique

75

that could be used to mass-produce various types of RSF sub-micron fibers with

76

potential applications in regenerative medicine. Here, to the author’s knowledge we

77

have for the first time employed SBS to process RSF into randomly aligned and aligned

78

non-porous micro/nano- fibers, utilizing both organic (formic acid) and aqueous solvent

79

systems, channeling the flow of air to the collector. Porous fibers from an aqueous

80

system were obtained by using a modified SBS system based on a thermally induced

81

phase-separation (TIPS) technique.1,37 A co-axial SBS system was also employed to

82

mimic a pH gradient as a means to replicate some of the physiological parameters of the

83

natural spinning process26 and drive secondary protein structure formation.

84

2. EXPERIMENTAL SECTION

85

2.1 Regenerated silk fibroin

86

RSF solutions from Bombyx mori silkworm cocoons (Wildfibres; Birmingham, UK)

87

were prepared following a previously described protocol

88

Briefly, cocoons were dewormed and degummed in a 0.02 M boiling Na2CO3 (Sigma-

38

35,36

,

with minor modifications.

4


Page 5 of 39 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biomacromolecules

89

Aldrich, UK) aqueous solution for 30 min to remove the sericin. Degummed silk fibroin

90

was thoroughly washed in deionized water (three water changes, 20 min each) and air-

91

dried for 24h. The dried degummed silk fibroin was dissolved in 7.9 M aqueous LiBr

92

(Sigma-Aldrich, UK) to form a 16% w/v solution, with gentle continuous stirring at

93

60°C for 4h. The solution was centrifuged (10,000 ×g, 20 min) to remove any residual

94

pupa and dialyzed (3.5kDa MWCO SnakeSkin, ThermoScientific, UK) against 5L of

95

deionized water at 4°C for 3 days, with regular water changes under continuous stirring.

96

The concentration of the RSF solution after water dialysis was ~5-6% w/v. The solution

97

was centrifuged again (10,000 ×g, 20 min), and either freeze-dried for 72h (VirTis

98

BenchTop Pro; New York, USA) or concentrated for further use, described below.

99

2.2 Spinning dope formulations

100

Formic acid dopes as precursors of non-porous fibers were prepared by dissolving

101

freeze-dried RSF in formic acid (95% v/v; Sigma-Aldrich, UK) at a concentration of 8-

102

20% w/v at room temperature and stirred for 2h prior to spinning.

103

Aqueous precursor dopes were prepared by means of reverse dialysis processes.

104

Aqueous formulations as precursors for non-porous fibers were prepared by

105

concentrating an RSF solution by reverse dialysis (10 kDa MWCO SnakeSkin dialysis

106

tubing, Thermo-Scientific, UK) against a 15% w/v polyethylene glycol (PEG 10,000

107

Mw; Sigma-Aldrich, UK) aqueous solution at 4°C. Note that some PEG likely diffused

108

into the RSF solution and may have acted as a spinning aid. Prior to spinning, a 3M

109

CaCl2 (Sigma-Aldrich, UK) solution was added into the RSF system to adjust the [Ca]2+

110

to 0.3 M (~pH 7.5, 9:1 v/v ratio SF/CaCl2). The role of CaCl2 is to promote RSF

111

conformation from random coil/α-helix to β-sheet.39–41 The pH was subsequently

112

adjusted to 10.5 with a 5M NaOH (Sigma-Aldrich, UK) solution, which was found to

113

make the dope stable during the spinning process and promote fiber formation. Aqueous 5



Page 6 of 39

114

formulations as precursors of porous fibers and biomimetic fibers were prepared by

115

concentrating an RSF solution by reverse dialysis (3.5 kDa MWCO SnakeSkin dialysis

116

tubing, Thermo-Scientific, UK) against a 20% w/v PEG (10,000 Mw; Sigma-Aldrich,

117

UK) aqueous solution at 4°C.

118

All solutions were stored in sealed glass vials at 4ºC and spun within 24h.

119

2.3 Rheological analysis of regenerated silk fibroin solutions

120

RSF formulations at increasing concentrations were loaded into an AR-G2 rheometer

121

(TA Systems; UK), operated with a parallel plate geometry (20 mm diameter and gap

122

set to 600 μm) at 25°C and with an environmental case attached to prevent solvent

123

evaporation. Each concentration (n=2 replicates of the same solution) was subjected to a

124

logarithmic steady shear rate increase from 0.01 to 1000 s-1.42

125

2.4 Fiber production by conventional, cryogenic and co-axial solution blow

126

spinning

127

All RSF dopes were spun via a custom made SBS set-up (Scheme 1), consisting of a

128

syringe pump, connective tubing, a concentric nozzle and a compressed air source

129

controlled by a pressure regulator. Solutions were injected to the inner nozzle of the

130

SBS head using a precision syringe driver pump (N-300, New Era Pump Systems Inc.;

131

USA). Compressed air was delivered to the outer nozzle via an oil-free and water-free

132

air compressor (Bambi VT150D; Italy). The internal diameter (ID) of the SBS inner

133

nozzle was 0.8 mm, the outer diameter (OD) was 1.8 mm; the outer nozzle had an ID of

134

2 mm and an OD of 3 mm. The inner nozzle was set to protrude by ≥ +2 mm from the

135

outer nozzle into a conical shaped end.

136

Random non-porous fibers from RSF formulations based on formic acid were spun

137

directly with a conventional SBS apparatus (Scheme 1 a-b): feed rate of 50-60 6


Page 7 of 39 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biomacromolecules

138

μL.min−1, air pressure of 30-40 psi (0.21-0.28 MPa) and working distance to the

139

collector of 20-30 cm. Production of random non-porous fibers from RSF aqueous

140

dopes was enhanced by spinning through a vapor immediately after the SBS nozzle.

141

Vapor was generated using different solvents (water, absolute ethanol or absolute

142

methanol) in a beaker on a hotplate (100-200°C) placed under the SBS nozzle. Ethanol

143

and methanol are known to promote β-sheet formation. Spinning parameters in this case

144

were: feed rate of 100-110 μL.min−1, air pressure of 30-60 psi (0.21-0.41 MPa) and

145

working distance of 30-40 cm to fiber collection.

146

Aligned non-porous fibers out of formic acid and aqueous systems were obtained by

147

placing a funnel between the SBS head and collector (Scheme 1c). This served to

148

channel the airflow and the fiber forming jet, aligning the fibers in flight prior to

149

collection. The same spinning parameters as previously stated for non-porous fibers

150

were used.

151

Porous fibers were obtained using RSF aqueous dopes based on a cryogenic SBS (cryo-

152

SBS) set-up (Scheme 1d).1,37 As-produced RSF aqueous formulations were sprayed into

153

fibers directly into a bath of liquid nitrogen. The feed rate was 100 μL.min−1 and the air

154

pressure 30 psi (0.21 MPa); a working distance of 15 cm and an angle of ~45° to the

155

cryogenic bath were selected to limit liquid nitrogen loss. The bath container was made

156

from an expanded-polystyrene foam box. During cryo-SBS, the level of liquid

157

nitrogen in the bath was constantly topped up to ensure that the working distance

158

remained ±2 cm of the target value. Frozen fibers were then collected into 50 mL

159

Falcon® tubes, transferred to lyophilization flasks (pre-cooled in liquid nitrogen) and

160

subsequently lyophilized (VirTis BenchTop Pro; New York, US). The external glass

161

portion of the lyophilization flask was kept under liquid nitrogen using the polystyrene

162

foam box until a vacuum of ~50 mTorr was reached, after which the outside of the flask 7



Page 8 of 39

163

was exposed to ambient conditions. Lyophilization was continued for a total of 72h to

164

ensure complete drying. a

Vapor phase Pressure gauge

Pressurized gas

b Air nozzle

Collector

SF solution

Nozzle

Pressurized air

Solution cone

Polymer jet Fiber

Solution pump

c

d

Pressure gauge

Pressure gauge Funnel

Nozzle

Cryogenic bath

Nozzle

e

Air nozzle Outer solution nozzle Inner solution nozzle SF solution

Pressurized air

165

f

Air nozzle Inner solution nozzle Outer solution nozzle

SF solution Outer solution Self-assembly

Pressurized air

Self-assembly

166

Scheme 1 Schematic representation of conventional, cryogenic and co-axial SBS

167

set-ups. (a-b) Conventional SBS set-up (with/without vapor phase). (c) Modified SBS

168

set-up to channel the airflow and RSF fiber jet for spinning aligned non-porous fibers.

169

(d) Cryo-SBS set-up for the production of porous fibers. (e-f) Co-axial SBS set-up to

170

induce a pH-ion gradient by (e) positive and (f) negative inner nozzle solution

171

protrusion.

172

For biomimetic fibers, a co-axial SBS head, consisting of two concentric solution

173

nozzles (inner solution nozzle, outer solution nozzle) within an outer pressurized air

174

nozzle, was employed to induce changes to the chemical environment at the point of the 8


Page 9 of 39 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biomacromolecules

175

solution cone (positive protrusion of inner solution nozzle, Scheme 1e) or prior to it

176

(negative protrusion of inner solution nozzle, Scheme 1f) by flowing a secondary

177

solution through the outer solution nozzle whilst flowing the RSF solution through the

178

inner solution nozzle. The outer solution nozzle was set to protrude ≥ 2 mm from the

179

outer pressurized air nozzle; the inner solution nozzle was set to protrude ± 2 mm

180

(positively or negatively) from the outer solution nozzle. A number of kosmotropic

181

phosphate solutions adjusted to various acidic pH values (pH 2-7) and a commonly used

182

coagulant in wet-spinning (ammonium sulfate)

183

aqueous RSF solutions into fibers. Two precision syringe driver pumps were used for

184

injecting the inner and outer solutions. Successful fiber spinning was achieved with the

185

following set parameters. For the positive nozzle protrusion, inner and outer solutions

186

were injected at feed rates 100 μL.min−1 and 150 μL.min−1, respectively; for the

187

negative nozzle protrusion, the inner and outer solutions were injected at feed rates of

188

190 μL.min−1 and 100 μL.min−1, respectively. Air pressure and distance to collector

189

were maintained in both cases at 40 psi (0.28 MPa) and 40 cm.

190

2.5 Insolubilization of solution blow spun regenerated silk fibroin fibers

191

Non-woven random fiber mats were immersed in 90% v/v ethanol (EtOH) or 90% v/v

192

methanol (MeOH) baths for 10 min to further induce protein secondary structure

193

transition from α helix/random coils into β-sheets. The mats were then dried in an oven

194

at 30°C for 24h.

195

2.6 Physico-mechanical characterization of the fibers

196

2.6.1 Fiber morphology

197

Fiber morphology was assessed using scanning electron microscopy (SEM) on a Hitachi

198

S300 N (USA) and on a PhenomPro Generation 5 (USA), and field emission scanning

199

electron microscopy on a Zeiss Supra 35VP FE-SEM (Germany). Briefly, fiber mats

22

was used to induce self-assembly of

9



Page 10 of 39

200

(1cm x 1cm) were mounted onto aluminium stubs with double-sided conductive carbon

201

tape and sputter-coated with gold/palladium (Gatan Model 682 Precision Etching

202

Coating System, USA) to an average thickness of 3 nm; images were captured at 2.5-5

203

kV. Average fiber diameter was determined using ImageJ (Fiji 1.28, USA) software, by

204

measuring a minimum of 50 fibers per sample type.

205

2.6.2 Protein structure quantification and crystallinity assessment

206

The conformational structure of non-woven RSF fiber mats was investigated by wide-

207

angle x-ray diffraction (WAXD; PANalytical X’Pert Pro, UK) with the following

208

parameters: copper line focus X-ray tube with Ni kβ absorber (0.02 mm; Kβ=1.392250

209

Å) and Kα radiation (Kα1=1.540598 Å, Kα2=1.544426 Å, Kα ratio 0.5, Kαav=1.541874

211

Å), diffraction angle of 5-45° and scanning rate of 2°.min−1. In addition, a Fourier

212

with an iD5 attenuated total reflectance (ATR) attachment (diamond crystal) was used to

213

examine the conformational changes of the secondary structure of RSF within the amide

214

I region. For each measurement, 32 scans were recorded with a resolution of 4 cm-1 and

215

wavenumbers from 500 to 6000 cm-1. The amide I region ranging from 1600 to 1700

216

cm-1 was baseline corrected and deconvoluted by applying a second derivative

217

algorithm, a signal smoothing algorithm and a curve-fitting algorithm based on Fourier

218

Self-Deconvolution (FSD) according to references

219

at 1595 cm−1 (side chains), 1610 cm−1 (side chains), 1620 cm−1 (intermolecular β-

220

sheets), 1630 cm−1 (intermolecular β-sheets), 1640 cm−1 (random coils), 1650 cm−1

221

(random coils), 1660 cm−1 (α helices), 1670 cm−1 (β-turns), 1680 cm−1 (β-turns), 1690

222

cm−1 (β-turns) and 1700 cm−1 (intramolecular β-sheets) following reference 45. Samples

223

were run in triplicate.

224

2.6.3 Mechanical testing

210

transform infrared spectrometer (FTIR Nicolet iS5; Thermo Scientific, UK) equipped

43,44

. The band positions were fixed

10


Page 11 of 39 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biomacromolecules

225

Uniaxial tensile testing of as-spun RSF mats (cardboard window frame dimensions of 5

226

mm x 5 mm) of aligned non-porous fibers was performed with a mechanical tester

227

(Instron 3344; Instron Ltd., USA). The ends of the fiber mats were embedded in epoxy

228

and taped to avoid stress concentration at the clamps during mounting to the tensile test

229

machine. Prior to testing, the sides of the window frame were cut through, enabling

230

direct loading of the sample. Samples (n=12 replicates) were subjected to uniaxial loads

231

at 2 mm.min−1, a 10N load cell was used. The strain at break and ultimate tensile

232

strength (UTS) were calculated from the maximum value of the stress-strain curve

233

before breakage, Young’s modulus from the slope of the linear portion of the curve, and

234

toughness from the area under the graph for each sample tested. The area occupied by

235

the fibers was calculated by measuring the area of the mat and subtracting the void

236

space of the fibers based on mass-density differences between the mat and single

237

fibers.46 Instruments were calibrated prior to use and all tests were performed at room

238

temperature.

239

2.6.4 Surface area analysis and fiber pore size

240

Brunauer–Emmett–Teller (BET) surface area measurements of porous fibers obtained

241

by cryo-SBS were made using a surface area and porosity analyzer (TriStar II,

242

Micromeritics, Neu-Purkersdorf, Austria). Prior to gas adsorption experiments,

243

impurities were removed via a degassing step (FlowPrep 060, Micromeritics).

244

Approximately 100 mg of sample was placed inside a glass chamber, purged with

245

nitrogen at ambient temperature overnight. Nitrogen adsorption isotherms were then

246

measured at 77 K.

247

The size of a minimum of 70 pores from SEM images of porous fibers was measured

248

using by ImageJ (Fiji 1.28, USA); measurements were performed in the horizontal

249

direction through the center of the pore. 11



Page 12 of 39

250

2.7 Data analysis

251

Quantitative data is presented as standard deviation (SD) of the mean values. Statistical

252

analysis of the results was performed using GraphPad Prism 7 (San Diego, USA).

253

3. RESULTS AND DISCUSSION

254

3.1 Non-porous RSF fibers obtained via SBS

255

Analysis of the viscoelastic properties of RSF formulations is of interest since solution

256

viscosity determines whether a droplet spray or continuous fibers are obtained, as it

257

affects initial droplet shape and stability of the jet trajectory during SBS.47 The viscosity

258

measurements of RSF formulations of formic acid and the aqueous system at increasing

259

concentrations are shown in Figure 1 a-b. Both systems exhibited slightly different

260

rheologies over the range of concentrations tested. For the case of RSF solutions in

261

formic acid, the viscosities were almost independent of the shear rate exhibiting

262

Newtonian-like behavior. This is consistent with work reported by Zhu et al. for RSF

263

formulations in formic acid below 20% w/v.48 Aqueous RSF formulations, on the other

264

hand, exhibited non-Newtonian shear-thinning-like behavior, more evident at high

265

solution concentrations. Shear-thinning at increasing shear rate is characteristic of

266

native silk dopes49 and suggests more entanglement of the chain networks in the

267

polymer solution,50,51 which should aid spinning by SBS.

268

The solutions over the range of different concentrations were spun via conventional

269

SBS. A systematic investigation found that non-porous fibers only formed for

270

concentrations between 10-18% w/v for solutions based on the formic acid system, and

271

between 25-30% w/v for aqueous solutions in the presence of a water/ethanol/methanol

272

vapor and pH-Ca2+ adjustment as detailed in the experimental section. Solutions lower

273

than 10% and above 18% w/v (formic acid system), or below 25% and above 30% w/v

274

(aqueous system), did not have suitable rheological properties for processing into fibers 12


Page 13 of 39

275

by SBS: either the nozzle got blocked frequently or the solutions resulted exclusively in

276

droplets and no fibers were formed. The spinnability windows (Figure 1 c-d) for RSF

277

solutions were established as 15-144 mPa.s (formic acid system) and 500-1500 mPa.s

278

(aqueous system), identified as the target viscosities for blow spinning RSF dopes into

279

non-porous fibers based on formic acid and aqueous systems respectively. a

b

102

Viscosity (Log[Pa.s])

Viscosity (Log[Pa.s])

101 100 10-1 10-2 10-3 100

101

102

102 101 100 10-1 10-2 10-3 100

103

101

Shear rate (Log[1/s])

c

102

103

Shear rate (Log[1/s])

d

2

5x10

3x103 2

4x10

Viscosity (mPa.s)

Viscosity (mPa.s)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biomacromolecules

3x102 2x102 1x102

2x103

1x103

0

0

5

280

281

10

15

20

5

SF concentration (% w/v)

10

15

20

25

30

35

SF concentration (% w/v)

5% w/v

7% w/v

8% w/v

10% w/v

11% w/v

12% w/v

15% w/v

16% w/v

17% w/v

20% w/v

21% w/v

25% w/v

27% w/v

32% w/v

282

Figure 1 Rheology of RSF solutions. (a-b) Shear sweep measurements of (a) formic

283

acid and (b) aqueous RSF solutions at increasing concentrations. Red arrows represent

284

increase in concentration. (c-d) Relationship between viscosity (at 1000 s-1 shear rate)

285

and concentration for (c) formic acid and (d) aqueous solutions, highlighting the

286

spinnability window for SBS; error bars in SD (n=2). 13



Page 14 of 39

287

SEM images of RSF non-porous fibers spun from formic acid are shown at 12 and 15%

288

w/v (Figure 2 a-b, Figure S1). The most stable fiber mats were obtained at a flux of 50-

289

60 μL.min−1 at an air pressure of 30-40 psi (0.21-0.28 MPa) with a working distance of

290

20-30 cm to fiber collection. Increasing the concentration of the RSF formulation

291

resulted in fibers with a greater diameter consistent with previously reported data of

292

solution blow spinning of other polymers.3,52,53 The mean diameter of randomly aligned

293

fibers based on formic acid (Figure 2a, Figure S1a) was 290 (±60) nm and 910 (±300)

294

nm for RSF solution concentrations at 12 and 15% w/v, respectively. Fiber morphology

295

was relatively smooth in both cases. Aligned fiber bundles and mats were also produced

296

at 12% w/v (Figure 2b, Figure S1b) by channeling the airflow towards the collector.

297

Fibers were quickly formed as the RSF solution exited the nozzle and consolidated

298

within less than 5 cm, meaning that by channeling the flow through greater distances

299

fiber alignment could be achieved. Aligned non-porous fibers exhibited a similar mean

300

fiber diameter (p-value non-significant) compared to that of randomly aligned non-

301

porous fibers produced from RSF formulations at the same concentration. The mean

302

fiber diameter in this case was 260 (±150) nm, and fiber morphology was smooth and

303

consistent, with no droplets/beads observed.

14


Page 15 of 39 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biomacromolecules

304 305

Figure 2 SEM images of RSF non-porous fibers from formic acid and aqueous

306

solutions. (a-b) Fibers spun from formic acid at 12% w/v: (a) randomly aligned fibers;

307

(b) aligned fiber bundle with inset at higher magnification. (c-d) Fibers spun from an

308

aqueous system at 27% w/v through a water vapor: (c) randomly aligned fibers; (d)

309

aligned fibers.

310

For aqueous RSF solutions, the most stable non-porous fibers were obtained at a flux of

311

100-110 μL.min−1 at an air pressure 30-60 psi (0.21-0.41 MPa) and working distance of

312

30-40 cm. SEM images of RSF fibers spun at 27 w/v % are shown in Figure 2 c-d. The

313

structure of the fiber mats consists of non-porous fibers interconnected into a 3D porous

314

non-woven network with some conglutination; with a mean diameter of 410 (±130) nm

315

for randomly aligned fibers (Figure 2c) and 465 (±170) nm for aligned fibers (Figure

316

2d) (p-value non-significant between them). Compared to RSF fibers produced from 15



Page 16 of 39

317

formic acid dopes (Figure 2 a-b), these fibers had poorer morphological homogeneity,

318

which can be attributed to differences in the solvent volatility. Water has a relatively

319

high boiling point and low volatility, resulting in residual moisture content prior to

320

collection.54 In contrast, formic acid is highly volatile, meaning fibers are readily

321

formed as the RSF jets through the SBS nozzle.

322

It should be noted that some low molecular weight PEG likely diffused into the aqueous

323

RSF solutions during reverse dialysis concentration, and that this acted as a spinning aid

324

during SBS promoting fiber formation. A plasticizing effect due to the presence of PEG

325

in polylactide (PLA) blends has been reported,55 considerably increasing the elongation

326

at break, while the tensile strength remained unchanged. Low molecular weight PEG is

327

difficult to spin on its own,56,57 and only its high molecular weight form, PEO, is

328

commonly used during electrospinning of aqueous biopolymer systems due to its higher

329

viscosity that aids stabilizing the polymer jet.

330

PEG is water-soluble and can be easily removed after fiber spinning. RSF solutions for

331

aqueous non-porous fiber spinning were also concentrated up to 30% w/v by forced air-

332

flow and by reverse dialysis in PEG using a 3.5 kDa membrane, yet none of the

333

prepared solutions were able to form fibers, suggesting that concentrations >30% w/v

334

may be required. However, this is disadvantageous for industrial processing due to the

335

easy transition of highly concentrated solutions into gel and clogging of the nozzle.

336

The structure of non-porous fibers was further characterized by ATR-FTIR and WAXD

337

(Figure 3 a-c). Indeed, the conformation and structure of proteins are an important

338

factor in both their physico-chemical and mechanical properties, which ultimately

339

affects their ability to interact with biological interfaces – an important consideration for

340

tissue engineering and therapeutic delivery applications. 16


Page 17 of 39 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biomacromolecules

341

The ATR-FTIR spectra of the amide I (1700-1600 cm−1) and amide II (1600-1450 cm−1)

342

bands of non-porous fibers spun from formic acid and aqueous solutions are shown in

343

Figure 3a, indicating that the typical chemical composition of silk fibroin remained

344

unaffected during SBS. A clear shift of the absorption peak of the amide I region (1700-

345

1600 cm−1) after post-treatment in ethanol and methanol can be observed.

346 347

A detailed quantification of the amide I band by curve fitting and deconvolution

348

revealed changes in the chain structure (Figure 3c). While β-sheets make up the

349

crystalline regions of silk fibroin, other more flexible structures such as β-turns or

350

random coils and α-helices account for the more amorphous structures of silk.58,59

351

Significantly greater (p

Porous, Aligned and Biomimetic Fibers of Regenerated Silk Fibroin

Recommend Documents