Accepted Manuscript

Modulation of the mechanical, physical and chemical properties of polyvinyli-dene fluoride scaffold via non-solvent induced phase separation process fornerve tissue engineering applications

Nadia Abzan, Mahshid Kharaziha, Sheyda Labbaf, Navid Saeidi

PII: S0014-3057(17)32337-6DOI: https://doi.org/10.1016/j.eurpolymj.2018.05.004Reference: EPJ 8402

To appear in: European Polymer Journal

Received Date: 10 January 2018Revised Date: 1 May 2018Accepted Date: 5 May 2018

Please cite this article as: Abzan, N., Kharaziha, M., Labbaf, S., Saeidi, N., Modulation of the mechanical, physicaland chemical properties of polyvinylidene fluoride scaffold via non-solvent induced phase separation process fornerve tissue engineering applications, European Polymer Journal (2018), doi: https://doi.org/10.1016/j.eurpolymj.2018.05.004

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customerswe are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, andreview of the resulting proof before it is published in its final form. Please note that during the production processerrors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

1

Modulation of the mechanical, physical and chemical properties of

polyvinylidene fluoride scaffold via non-solvent induced phase

separation process for nerve tissue engineering applications

Nadia Abzan, Mahshid Kharaziha*, Sheyda Labbaf, Navid Saeidi

Department of Materials Engineering, Isfahan University of Technology, Isfahan, 84156-83111, Iran

Abstract

The aim of this research was to develop microporous poly(vinylidene fluoride) (PVDF) scaffolds with an

intrinsic electrical property, via the combination of non-solvent induced phase separation (NIPS) and

thermal induced phase separation (TIPS) process referred as N-TIPS method. For this purpose, the effects

of non-solvent incorporation (distilled water) in the solvent composition (N,N-dimethylformamide

(DMF)), immersion time at coagulation bath (1, 3, 6 and 24 h) as well as coagulation bath temperature (-

10, 0 and 20°C) and composition (DMF:water volume ratio= 2:6 and 6:4) on the properties of the

produced scaffolds were investigated. Results confirmed that N-TIPS processing parameters had a

profound effect on the morphological, mechanical, physical and thermal properties of the PVDF

scaffolds. For instance, with increased bath temperature, the formation of three-dimensional bi-continuous

scaffold with average pore size of 4.2±0.6 m was achieved, whereas increased in the immersion time in

coagulation bath from 1 to 24 h induced cellular morphology with a larger pore size. The formation of a

relatively small pore size and uniform foam-like structure at 3h soaking in coagulation bath showed to

improved mechanical properties of the scaffolds. It was also found that toughness of the scaffolds

significantly promoted from 27.5±16.4 MPa (after 1 h soaking) to 155.2±25.4 MPa (after 3 h soaking).

Moreover, depending on the functional parameters of N-TIPS process, β phase fraction and crystallinity

of the PVDF scaffolds were in the range of 61-87% and 30-47%, respectively. Remarkably, 3 h soaking

of PVDF polymer solution in coagulation bath with composition of 6:4 (D-3h-64Wscaffold) significantly

2

enhanced the crystallinity (47.03%) and β phase fraction (87.9%) and reduced crystallite size of PVDF

polymer. The PC12 cell attachment and proliferation on PVDF scaffolds prepared at various parameters

were also investigated. Noticeably, D-3h-64 scaffold with enhanced crystallinity and β phase fraction and

significantly higher toughness could promote cell spreading and proliferation. The results presented in

this study show a great potential of PVDF scaffolds with desired properties for nerve tissue engineering

application.

Keywords: Phase separation, Poly(vinylidene fluoride), Peripheral nervous Peripheral nervous

regeneration, Three-dimensional scaffold.

1.Introduction

When peripheral nerve injury results in a gap larger than 5 mm in length, the main challenge is to find an

alternative solution to autograft transplantation due to the limitation of this procedure [1,2]. Although

nerve autograft is the first strategy in nerve reconstructions, lack of functional recovery, the probability of

neuroma formation and structural differences between damaged nerve and donor site leads to the

development of an artificial nerve tube [1,3,4]. Through a structural and mechanical simulation of the

native extracellular matrix (ECM) and inhibition of fibrous scar tissue formation, nerve guidance channels

provide a desirable environment for directing regenerated axons [5]. In addition, for a successful nerve

repair, artificial nerve grafts must be electrically conductive to enhance nerve regeneration [3,6].

However, external power source is required to induce electrical signals to the cells for nerve regeneration.

In this respect, piezoelectric materials have been suggested for nerve tissue engineering to apply electrical

stimulatory cues, using mechanoelecterical transduction, without external power supply and therefore

wires [3,7]. Piezoelectric materials create transient surface charges by mechanical deformations, without

any additional energy sources and have been shown to increase the regenerated axons and neurite

outgrowth [6].

3

In order to provide piezoelectric nerve conduits for peripheral nerve tissue engineering application,

various types of polymers and ceramics have been evaluated [8]. Among them, polyvinylidene fluoride

(PVDF) is a promising polymer due to its excellent properties such as thermal stability, good chemical

resistance and piezoelectric behaviour [9,10]. PVDF is a semi-crystalline polymer that has at least four

distinct crystalline phases due to the different chain conformations. The non-polar phase is and polar

phases are called and [11]. phase, the most piezoelectric responsive phase of PVDF, is usually

obtained from crystallization of PVDF, when solvent is removed at temperatures below 70 °C [12], or

from different strong solvents such as N,N-dimethyl formamide (DMF). Mechanical stretching of phase

film at temperatures between 70-90 °C is the most common way to obtain oriented phase [11,13].

PVDF has, hence, been widely applied for tissue engineering applications [3,12,14,7]. Lee et al. [6] used

PVDF–trifluoroethylene (PVDF–TrFE) to fabricate electrospun fibrous scaffolds. Their results

demonstrated the attachment of dorsal root ganglion neurons and extension of neurites on all fibrous

scaffolds. Similarly, Young et al. [15] designed microporous PVDF membranes by immersion-

precipitation method for nerve tissue engineering.

Various techniques have been applied to develop two and three dimensional scaffolds including

electrospinning [16], solvent casting-particulate leaching [17], immersion precipitation [18], melt molding

[19] and template-assisted synthesis and rapid prototyping (RP) technologies [20]. Among them,

thermally induced phase separation (TIPS) and non-solvent induced phase separation (NIPS) are fast,

controllable, and scalable approach for the fabrication of scaffolds with interconnected porous network

required for tissue engineering applications [21,22,23]. While TIPS approach is based on the removing

the thermal energy from the dope solution, the main driving force of NIPS method is based on solvent and

non-solvent interactions [24]. Based on the differences between the principles of the two techniques, the

morphology of the obtained scaffolds is different and depended on the dimensional of the scaffolds [20].

For instance, NIPS technique often leads to the production of scaffolds with macro-voids or finger-like

structures in their cross-section showing unsatisfactory mechanical strength, whereas TIPS approach

4

results in different structures, such as ‘‘cell-like’’ structure, ‘‘finger-like’’ structure and ‘‘sponge-like’’

structure [25,26]. Lee et al. [27] applied TIPS method to fabricate PLGA/poly(c-glutamic acid)/Pluronic

17R4 porous nerve conduits with bimodal open pores to stimulate growth of Schwann cells for nerve

regeneration. Based on their results, these nerve conduits with bimodal open pores were new generation

of scaffolds which could be applied for nerve regeneration [27]. According to the fundamentals of TIPS

and NIPS techniques, when both approach are utilized in scaffold fabrication it is referred to as N-TIPS.

Previous studies have shown that the combination of these two techniques yields highly permeable

membranes with strong mechanical properties [24].

To date, N-TIPS has mainly been applied to develop PVDF membranes for filtration purposes, while

the role of N-TIPS processing parameters on the crystallinity, mechanical properties and phase

formation of PVDF has not been studied. Therefore, the main aim of the current study was to develop

PVDF scaffold using N-TIPS approach by altering processing parameters (bath temperature, immersion

time and coagulation bath composition) to produce optimized scaffolds with physical, mechanical and

biological properties suitable for nerve tissue engineering applications.

2. Materials and methods

2.1. Materials

PVDF (Mw = 275000 gr/ml) and dimethyl sulfoxide (DMSO) were purchased from Sigma-Aldrich.

DMF, ethanol (EtOH) and hexane were provided from Merck Co, respectively. Dulbecco's Modified

Eagle Medium-high glucose (DMEM-HI), horse serum (HS), fetal bovine serum (FBS) and

penicillin/streptomycin (pen/strep) were all obtained from Bioidea, Iran.

2.2. Fabrication of PVDF scaffolds

PVDF porous scaffolds were fabricated via N-TIPS method, , as presented in Fig. 1. Briefly, 10 wt.%

PVDF solution was prepared in different ratios of solvent (DMF): non-solvent (distilled water)

composition (100:0 and 96:4) at 60 ᵒC for 2 h. The homogeneous solution was casted in a petri-dish pre-

5

treated using Teflon and then soaked in coagulation bath. The immersing time (1, 3, 6 and 24 h) as well as

the temperature (-10, 0 and 20 ᵒC) and composition (DMF: water volume ratio= 2:6 and 6:4) of the

coagulation bath were varied to study their effects on the scaffold morphology. Consequently, the

scaffolds were washed with EtOH for 24 h to extract any residual solvent and then washed with hexane

three times to remove residual EtOH. Finally, the scaffolds were freeze-dried (Alpha 1-2LDplus,

Germany) for 8 h. It needs to mention that, in order to provide uniform and morphology in the whole

experiments, the thickness of scaffolds was fixed to about 50 m. According to the working parameters

consisting of solvent composition, soaking time and coagulation bath compositions the samples were

named as presented in Table 1.

2.3. Characterization of PVDF scaffolds

The surface morphology and cross section of the scaffolds was evaluated using scanning electron

microscope (SEM, Philips, XL30). The pore size of the scaffolds was determined using Image J software.

The chemical properties of the scaffolds consisting of chemical functional groups, β phase content and

crystallinity were investigated using Fourier transform infrared spectroscopy (FTIR, Bruker tensor,

performed over a range of 400-2000 cm-1

and resolution of 2 cm-1

) and X-ray diffraction (XRD, X0 Pert

Pro X-ray diffractometer, Phillips, Netherlands, CuKa radiation (λ= 0.154 nm)) techniques. β phase

content fraction (F(β)) was estimated using FTIR spectra, according to the equation (1) [9]:

(1)

where Xα and Xβ are crystalline mass fractions of α and β, and Kα and Kβ are the absorption coefficient

of each phase. Moreover, Aα and Aβ correspond to their absorbance at 760 and 840 cm–1

, respectively. In

this equation, it was assumed that only α and β phases are present. The absorption coefficient of α and β

phases are 6.1104

cm2.mol

-1 and 7.710

4 cm

2.mol

-1, respectively [28]. Moreover, the crystallinity (Xc) of

PVDF scaffolds was assessed using XRD patterns, according to the equation (2) [9]:

cc

c a

IX

I I

(2)

6

where Ic and Iα are the integrated intensities scattered by the crystalline and the amorphous phases,

respectively [9,11]. Additionally, crystallite size of polymers was estimated using the XRD patterns and

according to Scherrer equation (equation (3)):

coscrist

kB

L

(3)

Differential scanning calorimetry (DSC, METTLER TOLEDO DSC 1) was carried out to estimate

the heat of fusion (∆Hf), melting temperature (Tm) and crystallinity of PVDF scaffolds prepared at various

conditions. DCS analysis was performed under nitrogen atmosphere at a heat-cool-heat temperature

program with a heating rate and cooling rate of 10°C min-1

and 5°C min-1

, respectively, over the

temperature range from 25 to 200°C. Moreover, crystallinity of the scaffolds was also estimated using

DSC results and according to the equation (4) [26]:

0

*100fc

f

HX

H

(4)

where ∆Hf and are the estimated fusion heat of sample at melting, and the fusion heat for 100%

crystalline PVDF polymer. value of PVDF was supposed to 104.5 J.g

-1 [1].

Mechanical properties of PVDF scaffolds were determined using a tensile tester (Hounsfield H25KS).

Prior to mechanical testing, the rectangular specimens with dimension of 20×10 mm and average

thickness of 50 m were immersed in water for 1 day. After plotting the stress-strain curves (n =3), the

mechanical properties consisting of strain at break (elongation), tensile strength and modulus were

calculated. The tensile modulus was calculated from the initial 0-3% of linear region of the stress-strain

curves.

2.4. Cell culture

In order to evaluate the effect of scaffold morphology on their biological behavior, PC12 cells (a rat

neuronal cell line, Pasteur Institute of Iran (NCBI code: C153)) were exposed to the scaffolds. Before cell

seeding, the scaffolds were sterilized for 30 min in 70% (v/v) ethanol and, then, 2 h under ultraviolent

(UV) light and subsequently, immersed in complete culture medium overnight. The PC12 cell-line was

cultured in DMEM-HI supplemented with 10% (v/v) HS, 5 (v/v) % FBS, and 1 (v/v) % pen/strep at 37 °C

in a humidified atmosphere containing 5% CO2. After 70-80% of confluency, the cells were counted and

seeded on scaffolds (n=3) and also tissue culture plate as a control with a density of 104 cells/well. Cells

7

were incubated for 1, 3 and 5 days at 37 °C under atmosphere consisting 5%CO2. In all experiments,

medium was changed every three days.

2.4.1. Cell spreading, viability and proliferation evaluation

Attachment and spreading of PC12 cells cultured on the scaffolds for 7 days was evaluated using

SEM technique. After 3 h fixation with 2.5 (v/v) % glutaraldehyde (Sigma), the samples were rinsed with

PBS, and dehydrated in the gradient concentrations of ethanol (30, 70, 90, 96 and 100 (v/v) % for 10 min,

respectively. Finally, they were air dried, gold-coated and evaluated using SEM imaging.

The viability of PC12 cells seeded on various scaffolds were investigated using 3-(4,5-

dimethylthiazolyl-2)-2,5-diphenyl tetrazolium bromide (MTT) assay. At the specific incubation times (1,

4 and 7 days), after discarding the culture medium, the samples and controls (n=3) were incubated with

MTT solution with the concentration of 0.5 mg/ml for 4h. After formation of formazan, DMSO was

added to each sample to dissolve stabilized crystals and kept for 1 h at 37°C. Finally, the optical density

(OD) of the samples was measured with microplate reader (BioTek, Model ELX800 Instruments) against

DMSO (blank) at a wavelength of 490 nm. The relative cell viability (% control) was calculated as below

[5]:

Where ASample, Ablank and Acontrol were absorbance of the sample, blank (DMSO) and control (TCP),

respectively. Moreover, Metabolic activity of PC12 was determined by using the Resazurin assay at days

1, 4 and 7 of culture on the scaffolds and tissue culture plate (TCP, control). The Resazurin Assay is

based on the reduction of Resazurin, a normally non-fluorescent compound, to resorufin, a fluorescent

metabolite, due to the highly reducing milieu of a living cell [29]. In this regard, after discarding the

culture medium, Resazurin solution (concentration of 10 μg/ml in complete medium) was added to each

sample and kept in incubator for 4 h, until the color of the Resazurin solution changed. Subsequently,

Relative cell survival (%control) = *100sample b

c b

A A

A A

(5)

8

after transferring to a 96-well plate, the absorbance was read at 490 nm using microplate reader. Finally,

the normalized metabolic activity with respect to day1 was calculated for each scaffold.

2.5. Mathematical modeling process

Nucleation and growth kinetics of crystallite is an important factor in determining the final structure

and pore size and thereby, mechanical properties of the PVDF scaffolds. In order to investigate the

growth (crystallization) kinetics of polymer-lean phase (or solvent crystallite) in phase separation process,

mathematical modeling process were performed. In this way, effect of different coagulation bath (DMF:

water volume ratio= 2:6 and 6:4) and solvent (DMF: water volume ratio= 96:4 and 100:0) on the

nucleation and growth behavior were studied, similar to previous researches [30]. Considering the

diffusional behavior of the crystallization (nucleation and growth) of the polymer-lean phase, the

following equation was proposed for the prediction of the kinetics of this behavior:

Dα=Kt (6)

where K is the growth rate of the polymer-lean phase (µm/h), α is a constant relating to the nucleation

mechanism of the polymer-lean phase, t is the immersing time (h) and D is average pore size of the

scaffolds (µm). This kind of equation was already used for other diffusional growth phenomena

[30,30,31].

2.6. Statistical analysis

Statistical analyses were performed using one-way ANOVA (n≥ 3) and reported as mean ± standard

deviation (SD). To determine a statistically significance difference between groups, Tukey’s post-hoc test

using GraphPad Prism Software (V.6) with a p-value <0.05 considered to be significant.

3. Results and Discussion

3.1. Morphological characterization of PVDF scaffolds: Effect of coagulation bath temperature

In order to provide acceptable microenvironment for nerve regeneration, the scaffolds should

structurally and mechanically mimic natural ECM of nerve tissue. In this way, three dimensional

9

scaffolds with pores in the range of 50 nm to 50 μm are favorable to use as nerve guidance channels [32].

This range of pore size within a scaffold enables the scaffolds being permeable to the entry of nutrients

and oxygen into the channel, while providing the necessary barrier to prevent the infiltration of unwanted

tissues into the scaffold from outside and exiting of Schwann cells and growth factors from inside to

outside of the channel [33].

Since N-TIPS process is driven by the solvent/non-solvent exchange, one of the effective ways to

control the structure of the scaffold is to modulate the exchange rate of solvent/non-solvent (diffusion

rate) by changing the composition and temperature of coagulation bath [24]. Primarily, the role of

temperature coagulation bath as the driving force of polymer crystallization in TIPS process is studied.

Fig. 2(A) presents the SEM images of PVDF scaffolds prepared in various temperatures using two

different solvent compositions (DMF: water volume ratio) of 100:0 and 96:4 at constant polymer

concentration, soaking time (24 h) and coagulation bath composition (DMF: water volume ratio=6:4). As

water was involved in coagulation bath, due to high miscibility of DMF in water, the effect of NIPS

process on the morphology of the scaffolds was unavoidable.

However, a dense surface layer, which is a common morphology of scaffolds prepared using NIPS

process, due to the fast solvent flow, was not detected and this could be due to the role of freeze drying

step in the formation of a porous scaffold. Therefore, all scaffolds revealed porous structure with various

pore sizes depending on the coagulation bath temperature and solvent type. Generally, at NIPS process,

when a stable polymer solution is exposed to the coagulation bath, it becomes thermodynamically

unstable leading to demixing (liquid-liquid ((L-L) phase separation) of solution and formation of a

polymer-rich and polymer-poor components. These two components finally transformed to the membrane

matrix and pores, respectively. Therefore, the demixing process and its rate significantly control the

physical properties of the PVDF scaffolds. The effects of coagulation bath temperature and composition

on the average pore size of the processed scaffolds are presented in Figs. 2(B-C).

10

With, increase in the bath temperature up to 20 °C, larger pores with bicontinuous structure at both

solvent were obtained. At DMF: water volume ratio of 100:0 (Fig. 2B), the average pore size of the

scaffolds increased significantly (P<0.05) from 0.7±0.1 m to 4.2±0.6 m with increased coagulation

temperature from -10 °C to 20 °C, which was suggested to be due to improved mass exchange rate at a

higher temperature (20 °C). When the solvent ratio was 96:4, while similar trend was detected, the pore

size changes were not noticeable. In this condition, the average pore size increased from 0.7±0.1 m to

2.4±0.7 m with increase in temperature (Fig. 2C). Moreover, according to the cross-sectional images of

the scaffolds prepared at -10 °C, high rate of solvent:non-solvent diffusion resulted in a macro-void

structure formation. A similar effect was previously reported for a PLLA/HA scaffold prepared at

coagulation bath with temperature of -18°C, 8°C and liquid nitrogen. It was found that the scaffolds made

at a quenching temperature of 8°C had a greater interconnectivity with larger pore sizes [34].

Furthermore, Lin et al. [34] fabricated polypropylene membranes with bicontinuous structure using TIPS

method. Their results showed that with increased quenching temperature (bath temperature) from 0 to 90

°C enhanced average pore size was detected. Jung et al. [24] also showed that increasing the temperature

of coagulation bath from 5 to 60 °C led to gradual disappearance of macrovoids in PVDF scaffold.

Nevertheless, based on the results of this study, low quenching temperatures of -10 and 0°C resulted in a

faster cooling rate and enabled a shorter time for nucleation and growth of solvent crystals and phase

separation, leading to the formation of smaller pores in scaffolds. Moreover, the presence of non-solvent

in the solvent component increased the crystallinity of the polymer and significantly reduces the average

pore size of the scaffolds from 4.2±0.6 μm to 2.4±0.7 μm). This might be due to the presence of water

crystals in the system that leaves smaller space for solvent crystals to nucleate and grow, leading smaller

pore formation [21]. According to our results, 20 °C was selected as the optimized coagulation

temperature for further characterization of N-TIPS approach

3.2. Morphological characterization of PVDF scaffolds: Effect of immersing time and solvent

composition

11

Fig. 3(A) shows SEM images of the surface and cross-section of PVDF scaffolds fabricated using

different solvents (100:0 and 96:4) and soaking times (1, 3, 6 and 24 h) at constant coagulation bath

temperature and composition. Moreover, Figs. 3(B-C) presents the effects of soaking time and solvent

composition on the average pore size of PVDF scaffolds. Due to similar coagulation bath temperature, the

driving force of TIPS approach remained unchangeable. Therefore, the modulation of the scaffold

morphology was primarily due to the changing of the NIPS parameters at different soaking times.

Results indicated that by increasing the soaking tome from 1 to 24 h, at solvent composition of 100:0,

all scaffolds reveal a bicontinuous, and uniform cellular structure (Fig. 3, cross-section images).

Moreover, the pore sizes of these scaffolds increased from 1.2±0.2 μm to 4.2±0.6 μm (Fig. 3(B)), with

increasing soaking time from 1 to 24 h. Similar to Fig. 2, when the solvent was 96:4, pore size of the

scaffolds reduced (0.9±0.2 μm-2.4±0.7 μm), depending on the soaking time, compared to scaffolds

prepared using 100:0 solvent. In this solvent composition, while 1 h soaking time in coagulation bath did

not result in a uniform cellular structure, increasing the soaking time induced the interconnected

morphology with foam-like structure. At solvent composition of 96:4, the presence of water in solvent

component, left smaller space for solvent crystals to nucleate and grow and so the average pore size of the

scaffolds decreased. Wei et al. [35] used dioxane/water mixture for solvent system to fabricate nano-

hydroxyapatite/PLLA composite scaffolds via TIPS method. They showed that the addition of small

amounts of water (5 wt.%) to solvent system (dioxane: H2O volume ratio=95:5) induced an

interconnected and random morphology with smaller pores that actually caused a reduction in mechanical

properties of the scaffolds. Moreover, it was found that the typical morphology of TIPS process

(spherulites) was not detected in the cross-section of the scaffolds (Fig. 3, cross-section images) which

could be helpful to control the mechanical properties of the scaffolds [25,24] . According to the previous

researches [37,38], spherulite structures could form from solid-liquid phase separation according to the

nucleation-growth mechanism of polymer crystals instead of passing across a bimodal line. Naturally,

membranes with spherulitic structures (disconnected) are weaker than membranes with a bicontinuous

(cellular) structure formed from a liquid-liquid phase separation [25].

12

SEM images of PVDF scaffolds fabricated in various soaking times and solvent compositions at

constant coagulation bath temperature (20°C) and composition (DMF: water volume ratio=2:6) is

presented in Fig. 4(A). After 1 h soaking in coagulation bath, the scaffolds revealed a dense surface with

macro-voids in their cross-sections. However, increasing immersing time up to 24 h induced much porous

structure with larger pore sizes (0.1±0.1-2.9±0.1 µm) (Figs. 4(B-C)). Moreover, cross-section SEM

images of scaffolds soaked for 3, 6 and 24 h showed a bicontinuous structure. According to our results, 1

h soaking in coagulation bath was not enough for external water to diffuse into the depth of the scaffolds

and provide a porous scaffold with a bicontinuous structure. After 3 and 6 h soaking in coagulation bath,

external water diffused into the depth of the scaffolds and induced more pores compared to the sample

prepared following 1 h soaking in coagulation bath. Finally, 24 h soaking in coagulation bath resulted in

the largest pore size for both solvent composition (2.9±0.1µm at D-24h-26 and 4.9±0.3 µm at DW-24h-

26). Moreover, the presence of water in the solvent system induced larger pores in the scaffolds. It could

be attributed to the higher volume of water in the coagulation bath, which extracted more internal solvent

from the scaffold (in comparison to 96:4 composition) and as a result, smaller pores had formed.

Deshmukh et al. [38] investigated the effect of ethanol: water ratio in coagulation bath on the morphology

of PVDF hollow fiber membranes. They found that increasing the amount of ethanol in the bath (10%-

50%) large macrovoids were by a short finger-like or sponge-like structure [38].

Based on SEM images of the samples prepared at two different coagulation bath compositions (Figs.

2 and 3), it was evident that coagulation type is the most important factor responsible for the progression

of liquid–liquid demixing or crystallization of the PVDF scaffold and, hence, their morphology. During

immersion precipitation, water as a durable non-solvent in the coagulation medium resulted in rapid

liquid–liquid demixing process and consequently, formation of finger-like voids which is not preferred for

tissue engineering application. However, incorporation of water, along with solvent in the coagulation

bath could postpone liquid–liquid demixing of polymers and ultimately led to the formation of sponge-

like scaffolds [39]. It could be clearly seen that by increasing solvent concentration in the coagulation

bath (6:4 vs 2:6), the typical NIPS morphology (macro-voids) which is clearly observed in D-1h-26 and

13

DW-1h-26 samples was replaced by the sponge-like scaffolds. This changing in the morphology of the

scaffolds might be due to the reduced concentration gradient of solvents leading to slower mass exchange

rate at high solvent concentration coagulation bath and consequently delayed demixing process. In this

condition, porous bicontinuous PVDF scaffold was developed.

3.3. Modeling of growth behavior of polymer-lean phase

Most studies related to TIPS and NIPS processes have focused on the application of phase diagram to

determine the thermodynamic stability of polymer systems. However, these phase diagrams disregard the

kinetic factors consisting of the rate of polymer crystallization and the solvent/non-solvent diffusion,

which are evaluated as essential factors in the fabrication of scaffolds and their final properties such as

their pore size [25]. The most important parameters affecting the pore size of the scaffolds prepared via

N-TIPS method are temperature and composition of coagulation bath and solvent. In this research, bath

temperature was supposed to be constant (20 °C). Therefore, the role of other parameters on the pore size

was considered. In order to find the constant values of K and α, logarithm of both sides of equation 6 was

estimated according to the following equation:

Ln(D) =

+

(7)

Finally, Ln(D) vs Ln(t) at different soaking times of 1, 3, 6 and 24 h were employed. Figs. 3(D) and

4(D) shows the average pore size changes of the scaffolds (Ln(D)) as a function of immersing time

(Ln(t)) in coagulation bath with two different solvents (100:0 and 96:4) and coagulation bath (DMF:water

volume ratio=6:4 and 2:6), respectively. The slope and width from origin of the diagram were showed

and

, respectively. The calculations were performed for all conditions and their results, including α

and K values, are presented in Table 2. Results illustrated that the experimental data fit well in kinetic

model. Therefore, considering the statement of the last paragraph of section 2, we could attribute the

calculated values of K to the growth rate and also the α values to the nucleation mechanisms.

14

Subsequently, these calculated values, for different production conditions, can be compared with each

other. Interesting, it was found that incorporation of more solvent in coagulation bath (DMF:water

volume ratio=6:4) resulted in significantly higher rate of liquid–liquid demixing and hence, the nucleation

mechanism of polymer-lean phase.

3.4. Mechanical properties of PVDF scaffolds: effect of immersing time and solvent composition

Effects of soaking time at coagulation bath and solvent on the mechanical properties of the PVDF

scaffolds prepared at constant coagulation bath temperature (20°C) and composition (DMF: water volume

ratio=6:4) are presented in Fig. 5. At both solvent (100:0 and 96:4), mechanical properties (tensile

strength, elastic modulus, elongation and toughness) were significantly enhanced with increasing soaking

time up to 3 h. For example, at solvent composition of 96:4, tensile strength, elastic modulus and

toughness of the scaffolds significantly increased from 0.3±0.1 MPa, 2.29±1.2 MPa and 27.5±16.4 KPa,

respectively (after 1 h soaking) to 0.9±0.1 MPa, 9.5±1.9 MPa and 155.2±25.4 KPa, respectively (after 3 h

soaking) (P<0.05). This increase could be due to the uniformity in the morphology and pore size of the

PVDF scaffolds prepared after 3 h soaking in coagulation bath (Fig. 2). Due to the formation of larger

pores at longer soaking time, mechanical properties of the scaffolds were significantly reduced. For

example, at 96:4 solvent, tensile strength, elastic modulus and toughness prominently decreased to

0.5±0.1 MPa, 8.2±1.4 MPa and 35.3±8.9 KPa, respectively, following 24 h soaking in coagulation bath

(P<0.05). At 100:0 solvent, similar behavior with different order was detected. In this condition, with

increased soaking time to 3 h, tensile strength, elastic modulus and toughness of the scaffolds

dramatically enhanced (about 4, 3 and 5 times, respectively). Moreover, the mechanical properties of the

scaffolds prepared using two different solvents were not significantly different (P>0.05).

PVDF porous scaffolds fabricated in this work, revealed comparable mechanical properties (tensile

strength, elastic modulus and toughness) to those of previous studies. Ji et al. [26] fabricated a porous

PVDF membrane via TIPS method, using dibutyl phthalate (DBP) and di(2-ethylhexyl) phthalate (DEHP)

as a mixed solvent system. The uniform sponge-like structure of membranes revealed the tensile strength

of 0.6-1.1 MPa and elongation at break of 11-25%, depending on the diluent mixtures. Su et al. [25] used

15

a mixture of -butyrolactone (-BA), cyclohexanone (CO) and DBP solvent system to develop PVDF

membranes via TIPS process. They showed that the tensile strength of PVDF scaffold was in the range of

0.22-2.08 MPa, depending on the solvent type. Young et al. [15] developed microporous PVDF

membranes grafted with poly(acrylic acid) (PAA) using isothermal immersion precipitation method, for

neural applications. They used dimethyl sulfoxide (DMSO) and water as solvent and non-solvent,

respectively. Results showed that tensile strength of the scaffolds immersed in coagulation bath with

various compositions (water: DMSO volume ratio=100:0, 60:40 and 70:30) was 1.77, 1.55 and 1.22 MPa,

respectively. Based on our results, between the selected working conditions (at coagulation bath

composition of 6:4), D-3h-64 and DW-3h-64 were nominated as optimized scaffolds in the terms of

morphological and mechanical properties.

Effect of soaking time and solvent composition at constant coagulation bath composition (2:6) on the

mechanical properties of the scaffolds are presented in Fig. 6. Similar to previous coagulation bath

condition (6:4), increasing the immersion time upon to 3 h resulted in enhanced mechanical behavior of

the scaffolds. For example, at 96:4 solvent, tensile strength, elastic modulus and toughness increased from

0.29±0.06 MPa, 1.6±0.3 MPa and 22.76±2.5 KPa (at 1 h soaking) to 0.92±0.3 MPa, 6.7±1.2 MPa and

72.8±21.4 KPa (at 3 h soaking). Moreover, at solvent composition of 100:0, tensile strength, elastic

modulus and toughness improved (about 1.5, 2.5 and 1.5 times, respectively) with increasing immersion

time to 3 h. It might be due the presence of macro-voids which could weaken the mechanical properties of

the scaffolds. Furthermore, mechanical properties of the scaffolds reduced with increasing soaking time

upon 24 h, in both solvent. Therefore, D-3h-26 and DW-3h-26 scaffolds were selected as optimized

samples in terms of morphological and mechanical properties under the mentioned condition (coagulation

bath temperature= 20°C and composition of DMF: water volume ratio=2:6) for further characterization.

Moreover, our results demonstrated that D-3h-26 and DW-3h-26 scaffolds revealed significantly different

mechanical properties compared to the scaffolds fabricated using previous coagulation bath (6:4). It might

be due to the more uniform porous structure of the scaffolds fabricated at using DMF: water volume

ratio=6:4.

16

It is important for peripheral nerve conduits to have appropriate mechanical properties such as tensile

strength and toughness in order to resist environmental pressures and mechanical forces without

collapsing during in vitro and in vivo conditions. In addition, nerve channels need to be flexible enough as

high rigidity could damage regenerated axons and surrounding tissues [32,42,43]. Previous studies have

shown that tensile strength of normal nerve and acellular nerve is about 2.7 MPa and 1.4 MPa,

respectively [42]. Our results confirmed that PVDF scaffolds fabricated by N-TIPS method in this work,

presented sufficient mechanical properties for nerve tissue engineering in comparison to other researches

[43].

3.5. Physical, chemical and thermal properties of PVDF scaffolds

The crystalline phases of PVDF scaffolds with optimized mechanical properties and morphology

(DW-3h-64, D-3h-64, DW-3h-26 and D-3h-26) were identified via FTIR spectroscopy (Fig. 7(A)).

Between various phases of PVDF, polar β phase is favored due to its largest piezoelectric, ferroelectric

and pyroelectric coefficients, as well as high dielectric constant [44]. FTIR spectra confirmed the

presence of both α and β phase of PVDF in the scaffolds. Absorbance at 612, 760, 795, 853 and 974 cm–1

were corresponded to α phase, while the absorbance bands at 470, 511, 840, 878 and 1279 cm–1

were

related to the β phase [9,47,48]. β phase fraction of the optimized scaffolds estimated according to the

equation (1), are recorded in Table 3. Our results showed that D-3h-64 scaffold revealed maximum

amount of β phase (87.9%). Similarly, Gregorio [12] reported that β phase of PVDF is fevered when

crystallization occur from DMF solution at temperatures below 70°C. Moreover, the presence of water in

solvent system reduced β phase fraction of scaffolds. As an example, β phase of D-3h-64 scaffold reduced

to 72.2% when water was added in the solvent composition (DW-3h-64 scaffold). Previous researches

revealed that processing parameters such as solvent type [47], solution crystallization temperature,

external forces and mixing with other polymers [48] influenced the crystalline-phase formation during the

fabrication of PVDF scaffolds. Between them, the solution crystallization temperature revealed

significant role in the β-phase formation of PVDF. While the crystallization temperature less than 70 °C

led to the predominant β -phase of PVDF, a mixture of α - and β -phase of PVDF was formed at 70-110

17

°C. Furthermore, the solution crystallization temperature above 110 °C resulted in the formation of a-

phase of PVDF as predominate phase [49]. In another study, results revealed that -phase of PVDF

polymer could be predominate when butyrolactone solvent was applied [47]. It is well known that β

phase of PVDF, with all trans planar zigzag conformation, has the most dense structure and thus, the

forms with the most pyro- and piezoelectric properties [50]. It is believed that in the presence of water,

polymer chains have more space and move freely, while no external force prevents them from closely

packing into a dense structure. Therefore, the scaffolds at solvent composition of 96:4 revealed lower

fraction of β phase compared to water-free composition. Similarly, Ma et al. [51] revealed that the

addition of tetrahydrofuran (THF) with lower dipole moment in to DMF solvent resulted in less

interactions between polymers and solvents hindering the β -crystal of PVDF.

In addition to water content, immersion time in coagulation bath may have significant role in the

phase composition of PVDF scaffold. Fig. 7(B) shows the effect of soaking time of scaffolds in the water-

free condition at constant coagulation bath composition (DMF: water volume ratio=6:4) and temperature

(20°C) on the FTIR spectra of PVDF scaffolds. Intensity of the β phase absorption peaks was much

higher at D-3h-64 sample than DW-3h-64 and reduced by increasing soaking time. According to these

FTIR spectra and equation 1, β phase fraction of each scaffold was calculated. Our results showed that

increasing soaking time from 3 to 6 and 24 h resulted in reduced β phase fraction from 87.9% at 3 h to

and 64.5% and 61.9% at 6 and 24 h, respectively.

In order to estimate the role of coagulation bath composition and soaking time on the crystallinity and

crystallite size of PVDF polymer, XRD patterns of PVDF scaffolds at different optimized conditions

(DW-3h-64, D-3h-64, DW-3h-26 and D-3h-26) were studied (Fig. 7(C)). The well-known diffraction

peaks of α phase of PVDF were appeared at 2θ = 18.6, 27.8 and 39° assigned to the lattice planes of

(020), (001) and (002), respectively. Moreover, β phase peaks of PVDF could be detected at 2θ= 20.7,

20.8, 36.6 and 56.1° which were appointed to the lattice planes of (200), (110), (101) and (221),

respectively [46,54,55]. XRD patterns were clearly showed that the intensity of α and β phase changed at

18

various coagulation bath composition and soaking time. Crystallinity and crystallite size of the scaffolds

were estimated from XRD patterns according to equations 2 and 3, respectively and are presented in

Table 3. The degree of crystallinity of the scaffolds were in the range of 30-47%, depending on the

fabrication parameters. This result was in agreement with previous studies [41,28]. Moreover, crystallite

size of the scaffolds reduced from 16.6 nm to 5.2 nm with increase in β fraction and crystallinity of the

scaffolds. Ma et al. [54], fabricated PVDF membrane via NIPS method, using N,N-dimethylformamide/γ-

Butyrolactone (γ-BL) as solvent and water as non-solvent. They estimated the crystallinity of the

membranes under different ratios of DMF/γ-BL and different polymer concentrations, in the range of 30-

38%. Results in this study showed that D-3h-64 scaffold, which had the highest amount of β phase from

FTIR spectra, revealed the highest crystallinity (47.0%) and the lowest crystallite size (5.2 nm). Salimi

and Yousefi [50] studied the crystallinity of two different grades of PVDF resin ( Kynar 720 and Hylar

MP10) in the shape of films fabricated by molding method. Results in this study indicated that the degree

of crystallinity of the films was in the range of 35.3-37.8% for Kynar 720 and 36-42.6% for Hylar MP10.

In addition, Hylar MP10 which had the higher crystallinity, revealed the greater β phase fraction (74%).

They believed that under high temperature of processing, polymer [50].

Effects of immersing time, at the water-free condition and constant coagulation bath composition

(DMF: water ratio of 6:4) and temperature (20°C) on crystallinity and crystallite size of the scaffolds

were also investigated. XRD patterns of the scaffolds immersed for various times (3, 6 and 24 h) in

coagulation bath are presented in Fig. 7(D). Increasing the immersion time resulted in reduction of the

intensity of β phase peak of (200)/(110) and induced lower crystallinity and higher crystallite size of the

scaffolds (Table 3). The crystallinity of the PVDF scaffold reduced from 47% to 32% and 29% with

increasing soaking time form 3 h to 6 and 24 h, respectively. Furthermore, the crystallite size of the

PVDF polymer enhanced from 5.2 nm to 14.3 nm with promoting soaking time in coagulation bath from

3 h to 24 h.

Effect of water content in coagulation bath of N-TIPS process on the thermal properties of the

scaffolds (DW-3h-64 and D-3h-64) was also studied. DSC thermograms of the scaffolds as well as their

19

extracted data are presented in Fig. 8. By heating the scaffolds from 25 °C to 200 °C, melting point was

found around 170°C at both scaffolds. When the scaffolds were cooled from 200 °C to 25 °C, the

crystallization peaks were detected around 123 °C and 160-170 °C, at both scaffolds. The second

crystallization peak of D-3h-64 scaffold at 160-170°C was divided into two peaks (or a shoulder appeared

around 160-170 °C). The peak at lower temperature (detected in the figure) was attributed to α phase and

the peak at higher temperature was related to phase [50]. It could be clearly found that the β phase

related peak disappeared at DW-3h-64 scaffold. Moreover, in agreement with previous XRD patterns and

FTIR spectra, in the presence of water, crystallinity of the scaffolds decreased. This behavior could be

due to the competitive correlation between the crystallization and L-L phase separation during N-TIPS

approach. When water was added in the coagulation bath, PVDF crystallization is prior to the L-L phase

separation. Therefore, rapid phase separation rate could suppress the chance of PVDF chain gathering in

to the crystal lattice in the polymer-rich phase, leading to a lower crystallinity and less phase formation.

The role of solvent concentration in coagulation bath was similarly reported in previous researches. Ma et

al. [54] revealed that at higher DMF concentration in coagulation bath composed of DMF and γ-BL,

PVDF chains could easily gather in the polymer-rich phase leading to a higher crystallinity of the PVDF

scaffold.

3.6. Biological properties of PVDF scaffolds

Cell attachment and spreading are important factors in determination of biocompatibility of the

scaffolds. The SEM images of PC12 cells seeded on optimized scaffolds (DW-3h-64, D-3h-64, DW-3h-

26 and D-3h-26) are presented in Fig. 9. Cells attached to the porous scaffolds with spherical /rounded or

spreading morphology, depending on the scaffold type (cells have been pointed by arrows). Results

showed that cell attachment and spreading on the scaffolds enhanced with increasing β phase fraction of

PVDF scaffolds. For instance, on the D-3h-64 scaffolds with the highest amount of β phase (Table 3), the

cells oriented and clustered on the scaffold in a longitudinal fashion. In contrast, on the DW-3h-64

scaffold, the cells distributed with rounded morphology.

20

In addition to cell morphology, role of various PVDF scaffolds on the proliferation of PC12 cells was

examined via MTT and Resazurin assays. According to MTT assay (Fig. 10(A)), PC12 cell proliferation

enhanced on all PVDF scaffolds from day 1 to day 7. Remarkably, proliferation of PC12 cells on the D-

3h-64 scaffold significantly enhanced from 77±5 (%control) at day 1 to 149±6 (% control) at day 7.

Moreover, the proliferation of PC12 cells on water-free condition (D-3h-64 and D-6h-64) was

significantly higher than other scaffolds (P<0.05). For example, after 7 days, the proliferation of PC12

cells was 1.5-fold more on the D-3h-64 scaffolds (149±6 (%control)) than DW-3h-26 (102±8 (% control))

(P<0.05). Metabolic activity of PC12 was determined by using the Resazurin assay on the scaffolds and

TCP as control (Figure 10(B)). This assay also confirmed that the proliferation of cells increased on

various scaffolds with increasing culture time. Moreover, metabolic activity of PC12 cells on D-3h-26

scaffolds was higher than others.

When the cells attach to the surface of biomaterials, a sequence of chemical and physical interactions

happens between them leading to modulation of extracellular matrix deposition as well as cell

proliferation and differentiation [55]. In this regard, cell attachment is a crucial process which could be

affected by various physiochemical properties of biomaterial substrate [58,59,60]. In the recent decades,

aiming to control cellular responses, wide researches have focused on the modifying the physiochemical

properties of the biomaterials-based scaffolds consisting of chemical composition, mechanical properties,

hydrophilicity, surface topography and charge as well as porosity to induce suitable cell responses

[58,61,62,63]. For instance, size and porosity degree of the scaffolds on the cell proliferation and

demonstrated that cell-type depended role of these structural properties [58,64]. Nunes-Pereira et al. [14]

studied proliferation of MC3T3-E1 and C2C12 cells on P(VDF-TrFE) scaffold. They showed that higher

pore size and porosity of P(VDF-TrFE) scaffold induced cell elongation, referred just by the C2C12

muscle cells. In another study, Kharaziha et al. [64] found that attachment, proliferation and

differentiation of neonatal rat cardiac fibroblasts as well as protein expression of cardiomyocyte depended

on mechanical properties of poly(glycerol sebacate) (PGS):gelatin nanofibrous scaffolds. Along with

these properties, numerous tissues and cells are responsive to electrical and/or electromechanical stimuli,

21

such as cardiac muscle [64], bone [65] and nerves [5]. It was found that electrical stimulation markedly

improved viability, alignment, and contractile activities of cardiomyocytes seeded on carbon nanotube

(CNT)-PGS: gelatin fibrous scaffolds [64]. Recently, Hoop et al. [66] demonstrated that piezoelectric

PVDF could enable creation of electrical charges on its surface through acoustic stimulation which

induced neuritogenesis of PC12 cells. They revealed that the piezoelectric stimulation effect on the

neurite generation in PC12 cells was similar to the ones produced by neuronal growth factors. Moreover,

Lee et al. [6] also investigated the role of piezoelectric scaffolds on neurite extension of primary neurons

and demonstrated the potential use of piezoelectric fibrous scaffold for neural regeneration. In this study,

the effects of mechanical properties and phase of PVDF scaffold on the cell responses were

investigated. Taken altogether, our findings revealed that instantaneously enhanced mechanical properties

and phase of PVDF scaffold with remarkable piezoelectric property along with their greater mechanical

properties significantly promoted attachment, spreading and proliferation of PC12 cell cultured on PVDF

scaffold.

4. Conclusion

In summary, the combination of non-solvent induced phase separation (NIPS) and thermal induced

phase separation (TIPS) effects was applied to fabricate piezoelectric PVDF scaffolds for nerve tissue

engineering. Moreover, the effects of different working parameters such as coagulation bath composition

and temperature, soaking time and solvent composition on the physical, mechanical and biological

properties of porous PVDF scaffolds were investigated. Results showed that the incorporation of water as

non-solvent in dope solution decreased β phase fraction of PVDF scaffolds. In water-free condition, the

scaffolds prepared after 3 h soaking in coagulation bath with composition of DMF: water volume ratio=

6:4(D-3h-64), was selected as the optimized scaffold. Moreover, due to liquid-liquid phase separation

happening prior to PVDF crystallization, D-3h-64 scaffold revealed the highest crystallinity degree (47%)

and β phase fraction of PVDF (87.9%). Finally, PC12 cell attachment and proliferation significantly

22

enhanced on PVDF scaffolds with higher amount of β phase (D-3h-64 scaffold) making them suitable to

use as nerve guidance channel.

References:

[1] P.-H. Wang, I.-L. Tseng, and S. Hsu, “Bioengineering approaches for guided peripheral nerve regeneration,” J. Med. Biol. Eng., vol. 31, no. 3, pp. 151–160, 2011.

[2] L. He et al., “Manufacture of PLGA multiple-channel conduits with precise hierarchical pore

architectures and in vitro/vivo evaluation for spinal cord injury,” Tissue Eng. Part C Methods, vol.

15, no. 2, pp. 243–255, 2009. [3] L. Ghasemi‐ Mobarakeh et al., “Application of conductive polymers, scaffolds and electrical

stimulation for nerve tissue engineering,” J. Tissue Eng. Regen. Med., vol. 5, no. 4, 2011.

[4] A. Subramanian, U. M. Krishnan, and S. Sethuraman, “Development of biomaterial scaffold for nerve tissue engineering: Biomaterial mediated neural regeneration,” J. Biomed. Sci., vol. 16, no.

1, p. 108, 2009.

[5] N. Golafshan, M. Kharaziha, and M. Fathi, “Tough and conductive hybrid graphene-PVA: Alginate fibrous scaffolds for engineering neural construct,” Carbon N. Y., vol. 111, pp. 752–763,

2017.

[6] Y.-S. Lee, G. Collins, and T. L. Arinzeh, “Neurite extension of primary neurons on electrospun

piezoelectric scaffolds,” Acta Biomater., vol. 7, no. 11, pp. 3877–3886, 2011. [7] C. Ribeiro, V. Sencadas, D. M. Correia, and S. Lanceros-Méndez, “Piezoelectric polymers as

biomaterials for tissue engineering applications,” Colloids Surfaces B Biointerfaces, vol. 136, pp.

46–55, 2015. [8] F. Yang, R. Murugan, S. Wang, and S. Ramakrishna, “Electrospinning of nano/micro scale poly

(L-lactic acid) aligned fibers and their potential in neural tissue engineering,” Biomaterials, vol.

26, no. 15, pp. 2603–2610, 2005. [9] C. Tsonos et al., “Multifunctional nanocomposites of poly (vinylidene fluoride) reinforced by

carbon nanotubes and magnetite nanoparticles.,” Express Polym. Lett., vol. 9, no. 12, 2015.

[10] Z. Cui, E. Drioli, and Y. M. Lee, “Recent progress in fluoropolymers for membranes,” Prog.

Polym. Sci., vol. 39, no. 1, pp. 164–198, 2014. [11] E. Ozkazanc, H. Y. Guney, S. Guner, and U. Abaci, “Morphological and dielectric properties of

barium chloride‐ filled poly (vinylidene fluoride) films,” Polym. Compos., vol. 31, no. 10, pp.

1782–1789, 2010. [12] R. Gregorio, “Determination of the α, β, and γ crystalline phases of poly (vinylidene fluoride)

films prepared at different conditions,” J. Appl. Polym. Sci., vol. 100, no. 4, pp. 3272–3279, 2006.

[13] R. Magalhães et al., “The role of solvent evaporation in the microstructure of electroactive β-poly

(vinylidene fluoride) membranes obtained by isothermal crystallization,” Soft Mater., vol. 9, no. 1, pp. 1–14, 2010.

[14] J. Nunes-Pereira et al., “Poly (vinylidene fluoride) and copolymers as porous membranes for

tissue engineering applications,” Polym. Test., vol. 44, pp. 234–241, 2015. [15] T.-H. Young, J.-N. Lu, D.-J. Lin, C.-L. Chang, H.-H. Chang, and L.-P. Cheng, “Immobilization of

l-lysine on dense and porous poly (vinylidene fluoride) surfaces for neuron culture,” Desalination,

vol. 234, no. 1–3, pp. 134–143, 2008. [16] S. M. Damaraju, S. Wu, M. Jaffe, and T. L. Arinzeh, “Structural changes in PVDF fibers due to

23

electrospinning and its effect on biological function,” Biomed. Mater., vol. 8, no. 4, p. 45007,

2013. [17] Q. Fu, M. N. Rahaman, F. Dogan, and B. S. Bal, “Freeze-cast hydroxyapatite scaffolds for bone

tissue engineering applications,” Biomed. Mater., vol. 3, no. 2, p. 25005, 2008.

[18] D.-J. Lin, H.-H. Chang, T.-C. Chen, Y.-C. Lee, and L.-P. Cheng, “Formation of porous poly

(vinylidene fluoride) membranes with symmetric or asymmetric morphology by immersion precipitation in the water/TEP/PVDF system,” Eur. Polym. J., vol. 42, no. 7, pp. 1581–1594,

2006.

[19] M. E. Gomes, A. S. Ribeiro, P. B. Malafaya, R. L. Reis, and A. M. Cunha, “A new approach based on injection moulding to produce biodegradable starch-based polymeric scaffolds: morphology,

mechanical and degradation behaviour,” Biomaterials, vol. 22, no. 9, pp. 883–889, 2001.

[20] D. M. Correia et al., “Strategies for the development of three dimensional scaffolds from piezoelectric poly (vinylidene fluoride),” Mater. Des., vol. 92, pp. 674–681, 2016.

[21] R. Akbarzadeh and A. Yousefi, “Effects of processing parameters in thermally induced phase

separation technique on porous architecture of scaffolds for bone tissue engineering,” J. Biomed.

Mater. Res. Part B Appl. Biomater., vol. 102, no. 6, pp. 1304–1315, 2014. [22] F. Yang, R. Murugan, S. Ramakrishna, X. Wang, Y.-X. Ma, and S. Wang, “Fabrication of nano-

structured porous PLLA scaffold intended for nerve tissue engineering,” Biomaterials, vol. 25, no.

10, pp. 1891–1900, 2004. [23] X. Wen and P. A. Tresco, “Fabrication and characterization of permeable degradable poly (DL-

lactide-co-glycolide)(PLGA) hollow fiber phase inversion membranes for use as nerve tract

guidance channels,” Biomaterials, vol. 27, no. 20, pp. 3800–3809, 2006. [24] J. T. Jung, J. F. Kim, H. H. Wang, E. di Nicolo, E. Drioli, and Y. M. Lee, “Understanding the non-

solvent induced phase separation (NIPS) effect during the fabrication of microporous PVDF

membranes via thermally induced phase separation (TIPS),” J. Memb. Sci., vol. 514, pp. 250–263,

2016. [25] Y. Su, C. Chen, Y. Li, and J. Li, “Preparation of PVDF membranes via TIPS method: the effect of

mixed diluents on membrane structure and mechanical property,” J. Macromol. Sci. Part A Pure

Appl. Chem., vol. 44, no. 3, pp. 305–313, 2007. [26] G.-L. Ji, B.-K. Zhu, Z.-Y. Cui, C.-F. Zhang, and Y.-Y. Xu, “PVDF porous matrix with controlled

microstructure prepared by TIPS process as polymer electrolyte for lithium ion battery,” Polymer

(Guildf)., vol. 48, no. 21, pp. 6415–6425, 2007.

[27] J.-H. Lee and Y.-J. Kim, “Effect of bimodal pore structure on the bioactivity of poly (lactic-co-glycolic acid)/poly (γ-glutamic acid)/Pluronic 17R4 nerve conduits,” J. Mater. Sci., vol. 52, no. 9,

pp. 4923–4933, 2017.

[28] P. Martins, A. C. Lopes, and S. Lanceros-Mendez, “Electroactive phases of poly (vinylidene fluoride): determination, processing and applications,” Prog. Polym. Sci., vol. 39, no. 4, pp. 683–

706, 2014.

[29] S. Anoopkumar-Dukie, J. B. Carey, T. Conere, E. O’sullivan, F. N. Van Pelt, and A. Allshire, “Resazurin assay of radiation response in cultured cells,” Br. J. Radiol., vol. 78, no. 934, pp. 945–

947, 2005.

[30] C. Yue, L. Zhang, S. Liao, and H. Gao, “Kinetic analysis of the austenite grain growth in GCr15

steel,” J. Mater. Eng. Perform., vol. 19, no. 1, pp. 112–115, 2010. [31] J. MORAVEC, J. BRADÁČ, H. NEUMANN, and I. NOVÁKOVÁ, “Grain Size Prediction of

Steel P92 by Help of Numerical Simulations,” Met. Brno, pp. 508–513, 2013.

[32] C. Cunha, S. Panseri, and S. Antonini, “Emerging nanotechnology approaches in tissue engineering for peripheral nerve regeneration,” Nanomedicine Nanotechnology, Biol. Med., vol. 7,

no. 1, pp. 50–59, 2011.

[33] S. H. Oh and J. H. Lee, “Fabrication and characterization of hydrophilized porous PLGA nerve guide conduits by a modified immersion precipitation method,” J. Biomed. Mater. Res. Part A,

vol. 80, no. 3, pp. 530–538, 2007.

24

[34] R. Zhang and P. X. Ma, “Poly (α-hydroxyl acids)/hydroxyapatite porous composites for bone-

tissue engineering. I. Preparation and morphology,” 1999. [35] G. Wei and P. X. Ma, “Structure and properties of nano-hydroxyapatite/polymer composite

scaffolds for bone tissue engineering,” Biomaterials, vol. 25, no. 19, pp. 4749–4757, 2004.

[36] F. J. Hua, J. Do Nam, and D. S. Lee, “Preparation of a macroporous poly (L‐ lactide) scaffold by

liquid‐ liquid phase separation of a PLLA/1, 4‐ Dioxane/Water ternary system in the presence of NaCl,” Macromol. Rapid Commun., vol. 22, no. 13, pp. 1053–1057, 2001.

[37] H. K. Lee, A. S. Myerson, and K. Levon, “Nonequilibrium liquid-liquid phase separation in

crystallizable polymer solutions,” Macromolecules, vol. 25, no. 15, pp. 4002–4010, 1992. [38] S. P. Deshmukh and K. Li, “Effect of ethanol composition in water coagulation bath on

morphology of PVDF hollow fibre membranes,” J. Memb. Sci., vol. 150, no. 1, pp. 75–85, 1998.

[39] F. Liu, N. A. Hashim, Y. Liu, M. R. M. Abed, and K. Li, “Progress in the production and modification of PVDF membranes,” J. Memb. Sci., vol. 375, no. 1–2, pp. 1–27, 2011.

[40] G. H. Borschel, K. F. Kia, W. M. Kuzon, and R. G. Dennis, “Mechanical properties of acellular

peripheral nerve,” J. Surg. Res., vol. 114, no. 2, pp. 133–139, 2003.

[41] S. Ichihara, Y. Inada, and T. Nakamura, “Artificial nerve tubes and their application for repair of peripheral nerve injury: an update of current concepts,” Injury, vol. 39, pp. 29–39, 2008.

[42] D. Yucel, G. T. Kose, and V. Hasirci, “Polyester based nerve guidance conduit design,”

Biomaterials, vol. 31, no. 7, pp. 1596–1603, 2010. [43] N. Golafshan, H. Gharibi, M. Kharaziha, and M. Fathi, “A facile one-step strategy for

development of a double network fibrous scaffold for nerve tissue engineering,” Biofabrication,

vol. 9, no. 2, p. 25008, 2017. [44] D. M. Esterly, “Manufacturing of Poly (vinylidene fluoride) and Evaluation of its Mechanical

Properties.” Virginia Tech, 2002.

[45] J. Yang, X. Wang, Y. Tian, Y. Lin, and F. Tian, “Morphologies and crystalline forms of

polyvinylidene fluoride membranes prepared in different diluents by thermally induced phase separation,” J. Polym. Sci. Part B Polym. Phys., vol. 48, no. 23, pp. 2468–2475, 2010.

[46] T. Soulestin, V. Ladmiral, F. D. Dos Santos, and B. Améduri, “Vinylidene fluoride-and

trifluoroethylene-containing fluorinated electroactive copolymers. How does chemistry impact properties?,” Prog. Polym. Sci., vol. 72, pp. 16–60, 2017.

[47] M. Tazaki, R. Wada, M. O. Abe, and T. Homma, “Crystallization and gelation of poly (vinylidene

fluoride) in organic solvents,” J. Appl. Polym. Sci., vol. 65, no. 8, pp. 1517–1524, 1997.

[48] W.-K. Lee and C.-S. Ha, “Miscibility and surface crystal morphology of blends containing poly (vinylidene fluoride) by atomic force microscopy,” Polymer (Guildf)., vol. 39, no. 26, pp. 7131–

7134, 1998.

[49] R. Gregorio Jr and M. Cestari, “Effect of crystallization temperature on the crystalline phase content and morphology of poly (vinylidene fluoride),” J. Polym. Sci. Part B Polym. Phys., vol.

32, no. 5, pp. 859–870, 1994.

[50] A. Salimi and A. A. Yousefi, “Analysis method: FTIR studies of β-phase crystal formation in stretched PVDF films,” Polym. Test., vol. 22, no. 6, pp. 699–704, 2003.

[51] W. Ma, J. Zhang, S. Chen, and X. Wang, “β-Phase of poly (vinylidene fluoride) formation in poly

(vinylidene fluoride)/poly (methyl methacrylate) blend from solutions,” Appl. Surf. Sci., vol. 254,

no. 17, pp. 5635–5642, 2008. [52] A. Hartono, S. Satira, M. Djamal, R. Ramli, B. Herman, and E. Sanjaya, “Effect of mechanical

treatment temperature on electrical properties and crystallite size of PVDF Film,” Adv. Mater.

Phys. Chem., vol. 3, no. 1, p. 71, 2013. [53] S. Rajabzadeh, T. Maruyama, Y. Ohmukai, T. Sotani, and H. Matsuyama, “Preparation of

PVDF/PMMA blend hollow fiber membrane via thermally induced phase separation (TIPS)

method,” Sep. Purif. Technol., vol. 66, no. 1, pp. 76–83, 2009. [54] W. Ma, Y. Cao, F. Gong, C. Liu, G. Tao, and X. Wang, “Poly (vinylidene fluoride) membranes

prepared via nonsolvent induced phase separation combined with the gelation,” Colloids Surfaces

25

A Physicochem. Eng. Asp., vol. 479, pp. 25–34, 2015.

[55] C. Ribeiro, D. M. Correia, S. Ribeiro, V. Sencadas, G. Botelho, and S. Lanceros‐ Méndez, “Piezoelectric poly (vinylidene fluoride) microstructure and poling state in active tissue

engineering,” Eng. Life Sci., vol. 15, no. 4, pp. 351–356, 2015.

[56] H. Chen, Y. Liu, Z. Jiang, W. Chen, Y. Yu, and Q. Hu, “Cell–scaffold interaction within

engineered tissue,” Exp. Cell Res., vol. 323, no. 2, pp. 346–351, 2014. [57] P. X. Ma, “Biomimetic materials for tissue engineering,” Adv. Drug Deliv. Rev., vol. 60, no. 2, pp.

184–198, 2008.

[58] H. Shin, S. Jo, and A. G. Mikos, “Biomimetic materials for tissue engineering,” Biomaterials, vol. 24, no. 24, pp. 4353–4364, 2003.

[59] P. M. Martins et al., “Effect of poling state and morphology of piezoelectric poly (vinylidene

fluoride) membranes for skeletal muscle tissue engineering,” Rsc Adv., vol. 3, no. 39, pp. 17938–17944, 2013.

[60] H.-I. Chang and Y. Wang, “Cell responses to surface and architecture of tissue engineering

scaffolds,” in Regenerative medicine and tissue engineering-cells and biomaterials, InTech, 2011.

[61] B. J. Papenburg, E. D. Rodrigues, M. Wessling, and D. Stamatialis, “Insights into the role of material surface topography and wettability on cell-material interactions,” Soft Matter, vol. 6, no.

18, pp. 4377–4388, 2010.

[62] C. M. Murphy, M. G. Haugh, and F. J. O’Brien, “The effect of mean pore size on cell attachment, proliferation and migration in collagen–glycosaminoglycan scaffolds for bone tissue engineering,”

Biomaterials, vol. 31, no. 3, pp. 461–466, 2010.

[63] S. Rajabi-Zeleti et al., “The behavior of cardiac progenitor cells on macroporous pericardium-derived scaffolds,” Biomaterials, vol. 35, no. 3, pp. 970–982, 2014.

[64] M. Kharaziha et al., “PGS: Gelatin nanofibrous scaffolds with tunable mechanical and structural

properties for engineering cardiac tissues,” Biomaterials, vol. 34, no. 27, pp. 6355–6366, 2013.

[65] M. Kharaziha et al., “Tough and flexible CNT–polymeric hybrid scaffolds for engineering cardiac constructs,” Biomaterials, vol. 35, no. 26, pp. 7346–7354, 2014.

[66] D. M. Ciombor and R. K. Aaron, “Influence of electromagnetic fields on endochondral bone

formation,” J. Cell. Biochem., vol. 52, no. 1, pp. 37–41, 1993.

26

Figure captions:

Fig. 1. The schematic showing the N-TIPs approach followed by freeze drying method to

develop porous PVDF scaffold.

Fig. 2. Effect of coagulation bath temperature on the structural properties of PVDF scaffolds

prepared at constant coagulation bath composition (DMF:water volume ratio= 6:4) and after 24h

soaking: (A) SEM images as well as pore size of PVDF scaffolds at different bath compositions;

DMF: water volume ratio= (B) 100:0 and (C) 96:4 (*:P<0.05).

Fig. 3. Effects of soaking time on the microstructure of PVDF scaffolds prepared at constant

coagulation bath composition (DMF:water volume ratio= 6:4) and 20 °C. (A) SEM images

(surface and cross section images) and pore size of the scaffolds at different coagulation bath

compositions of DMF: water volume ratio= (B) 100:0 and (C) 96:4. (D)The relationship between

logarithmic pore size of the scaffolds and soaking time (t) at coagulation bath with two different

compositions (DMF: water =100:0 and 96:4) (*:P<0.05).

Fig. 4. Effects of soaking time on the microstructure of PVDF scaffolds prepared at constant

coagulation bath composition (DMF:water volume ratio= 2:6) and 20 °C. (A) SEM images

(surface and cross section images) (A) and pore size of the scaffolds at different coagulation bath

compositions of DMF: water volume ratio= (B) 100:0 and (C) 96:4. (D)The relationship between

logarithmic pore size of the scaffolds and soaking time (t) at coagulation bath with two different

compositions (DMF: water =100:0 and 96:4) (*:P<0.05).

Fig. 5. Mechanical properties of the PVDF scaffolds prepared at constant coagulation bath

composition (DMF:water volume ratio= 6:4) and 20 °C: Representative stress-strain curves of

the scaffolds prepared at (A) constant solvent composition (DMF:water=100:0) (A) and (B)

constant solvent composition (DMF:water=96:4) for various soaking times. Average (C)tensile

strength, (D) elastic modulus, (E) elongation and (F) toughness of the scaffolds prepared at

various conditions (*:P<0.05).

Fig. 6. Mechanical properties of the PVDF scaffolds prepared at constant coagulation bath

composition (DMF:water volume ratio= 2:6) and 20 °C: Representative stress-strain curves of

the scaffolds prepared at (A) constant solvent composition (DMF:water=100:0) and (B) constant

solvent composition (DMF:water=96:4) for various soaking times. Average (C) tensile strength,

27

(D) elastic modulus, (E) elongation and (F) toughness of the scaffolds prepared at various

conditions(*:P<0.05).

Fig. 7. FTIR spectra of the PVDF scaffolds (A) prepared at various solvent and coagulation bath

composition and (B) prepared at various soaking time in constant coagulation bath composition

(DMF:water volume ratio= 6:4) and temperature (20°C). XRD patterns of the PVDF scaffolds

(C) prepared at various solvent and coagulation bath composition and (D) prepared at various

soaking time in constant coagulation bath composition (DMF:water volume ratio= 6:4) and

temperature (20°C).

Fig. 8. (A) DSC thermograms and (B) their extracted data of D-20T-3h-6D4W and DW-20T-3h-

6D4W scaffolds.

Fig. 9. PC12 cell attachment on the different PVDF scaffolds; SEM images of PC12 cells after 7

days of culture on (A) D-3h-6D4W, (B) D-3h-2D6W, (C)DW-3h-6D4W and (D)DW-3h-2D6W

scaffolds.

Fig. 10. Cell viability of PC12 on various PVDF scaffolds measured using (A)MTT assays after

1, 4 and 7 days of culture and (B) Resazurin assay after 4 and 7 days of culture (data was

normalized according to day 1) (*:P<0.05).

28

Table 1: The list of samples based on the working parameters.

Sample Solvent composition

(DMF: water volume ratio) Coagulation bath composition

(DMF: water volume ratio) Soaking time (h)

D-1h-64 100:0 6:4 1

D-3h-64 100:0 6:4 3

D-6h-64 100:0 6:4 6

D-24h-64 100:0 6:4 24

D-1h-26 100:0 2:6 1

D-3h-26 100:0 2:6 3

D-6h-26 100:0 2:6 6

D-24h-26 100:0 2:6 24

DW-1h-64 96:4 6:4 1

DW-3h-64 96:4 6:4 3

DW-6h-64 96:4 6:4 6

DW-24h-64 96:4 6:4 24

DW-1h-26 96:4 2:6 1

Dw-3h-26 96:4 2:6 3

Dw-6h-26 96:4 2:6 6

DW-24h-26 96:4 2:6 24

29

Table 2: The extracted data from equation 7, including α and K values.

Sample K α

D-64 0.964 2.461

DW-64 0.543 3.512

D-26 0.311 1.858

DW-26 0.326 1.422

30

Table 3: Crystalinity, crystallite size and β phase fraction of PVDF scaffolds.

β phase(%) Crystallinity (%) Crystallite size

(nm) sample

87.9 47.0 5.2 D-3h-64

64.5 31.6 7.8 D-6h-64

61.9 29.4 14.3 D-24h-64

78.8 42.3 9.4 D-3h-26

61.1 30.4 16.6 DW-3h-26

72.2 36.4 13.1 DW-3h-64

31

32

33

34

35

36

37

38

39

40

41

Highlights:

PVDF scaffold was developed via non-solvent and thermal induced phase separation (N-TIPS).

Toughness of scaffolds promoted (5 times) with increasing soaking time in coagulation bath.

β phase of PVDF scaffolds changed in the range of 44-87% depending on the immersing time

Crystallinity of PVDF changed in the range of 30.35-47.03% depending on the immersing time.

PC12 proliferation on PVDF scaffolds significantly promoted via enhanced β phase fraction.

42