Static-Aligned Piezoelectric Poly (Vinylidene Fluoride) Electrospun Nanofibers/MWCNT Composite Membrane: Facile Method

Polyvinylidene Fluoride (PVDF) piezoelectric electrospun nanofibers have been intensively used for sensing and actuation applications in the last decade. However, in most cases, random PVDF piezoelectric nanofiber mats have moderate piezoelectric response compared to aligned PVDF nanofibers. In this work, we demonstrate the effect of alignment conducted by a collector setup composed of two-metal bars with gab inside where the aligned fiber can be formed. That is what we called static aligned nanofibers, which is distinct from the dynamic traditional technique using a high speed rotating drum. The two-bar system shows a superior alignment degree for the PVDF nanofibers. Also, the effect of added carbon nanotubes (CNTs) of different concentrations to PVDF nanofibers is studied to observe the enhancement of piezoelectric response of PVDF nanofibers. Improvement of β-phase content of aligned (PVDF) nanofibers, as compared to randomly orientated fibers, is achieved. Significant change in the piezoelectricity of PVDF fiber is produced with added CNTs with saturation response in the case of 0.3 wt % doping of CNTs, and piezoelectric sensitivity of 73.8 mV/g with applied masses down to 100 g.


Introduction
Piezoelectric polymer is a highly flexible type of polymer which can be deformed by stretching, compression, or bending, thereby producing an electrical polarization in response to applied stress. The polarization is caused by a reorientation of the net dipole moment of polymer polar groups into a particular direction [1][2][3][4][5]. Several types of semicrystalline polymers such as PVDF, Polyamides [6], Polyureas, and liquid crystal have been used for piezoelectric applications due to their high flexibility, strength, and impact resistance. Among different piezoelectric polymers, PVDF and its copolymer with tetraflouoroethylene (TFE) and trifluoroethylene (TrFE) represent the state of the art of piezoelectric a simple setup, which applied weights to a sandwich of PVDF nanofiber mats covered by two foil thin sheets. Then, the generated voltage was detected through a high impedance oscilloscope.

Materials
The main precursor of Polyvinylidene Fluoride (PVDF) (Kynar®, King of Prussia, PA, USA) is supplied by ARKEMA. Multi-walled carbon nanotubes (CNTs), from Cheaptubes Inc. (Cambridgeport, VT, USA) are added within different weight percentage. The outer diameter of these tubes is in the range of 10-20 nm, and the inner diameter between 3-5 nm.

Nanofiber Formation
Electrospinning was performed by adding 20 mL of dimethyl formamide (DMF) (Fine-Chem Limited, industrial Estate, Mumbai, India) solvent into a 3 gm polyvinylidene difluoride (PVDF) to get 15 wt % PVDF concentration. A plastic syringe tipped with a stainless steel needle was filled with 3 mL of the PVDF solution. The positive voltages came from a high voltage supply CZE1000R (Spellman, Hauppauge, NY, USA) to the metal needle, for application of bias values around (25 kV) with constant rate of (1.5 mL/h) using a syringe pump NE1000 (New Era Pump Systems, Suffolk County, NY, USA) with needle-to-collector distance of 10 cm. Random PVDF nanofibers were obtained using a normal metal plate collector covered with aluminum foil which was connected to ground. For comparison, Aligned PVDF fiber was fabricated by 2-metal bars of 1.5 cm length, as shown in Figure 1. The addition of MWCNTs with different concentration (0.1 and 0.3 wt %) was introduced by dispersing the CNTs into PVDF solution with the aid of tip sonicator (Fisher Scientific, Hampton, NH, USA) for 15 min in an ice bath to diminish the effect of heat on the CNT structure. through a simple setup, which applied weights to a sandwich of PVDF nanofiber mats covered by two foil thin sheets. Then, the generated voltage was detected through a high impedance oscilloscope.

Materials
The main precursor of Polyvinylidene Fluoride (PVDF) (Kynar®, King of Prussia, PA, USA) is supplied by ARKEMA. Multi-walled carbon nanotubes (CNTs), from Cheaptubes Inc. (Cambridgeport, VT, USA) are added within different weight percentage. The outer diameter of these tubes is in the range of 10-20 nm, and the inner diameter between 3-5 nm.

Nanofiber Formation
Electrospinning was performed by adding 20 mL of dimethyl formamide (DMF) (Fine-Chem Limited, industrial Estate, Mumbai, India) solvent into a 3 gm polyvinylidene difluoride (PVDF) to get 15 wt % PVDF concentration. A plastic syringe tipped with a stainless steel needle was filled with 3 mL of the PVDF solution. The positive voltages came from a high voltage supply CZE1000R (Spellman, Hauppauge, NY, USA) to the metal needle, for application of bias values around (25 kV) with constant rate of (1.5 mL/h) using a syringe pump NE1000 (New Era Pump Systems, Suffolk County, NY, USA) with needle-to-collector distance of 10 cm. Random PVDF nanofibers were obtained using a normal metal plate collector covered with aluminum foil which was connected to ground. For comparison, Aligned PVDF fiber was fabricated by 2-metal bars of 1.5 cm length, as shown in Figure 1. The addition of MWCNTs with different concentration (0.1 and 0.3 wt %) was introduced by dispersing the CNTs into PVDF solution with the aid of tip sonicator (Fisher Scientific, Hampton, NH, USA) for 15 min in an ice bath to diminish the effect of heat on the CNT structure.

Electric Field and Orientation Analysis
Nanofibers are formed between the 2-bars due to the high electric field. We introduced a finite element simulation for the proposed collectors during the study, showing both electric and electric field vectors between the 2 bars. The finite element package COMSOL Multiphysics® Version 5 (COMSOL Inc, Stockholm, Sweden) was used to demonstrate the fields profiles. The simulations were investigated using 3D module. SEM images were analyzed using Fourier Transform through the usage of MATLAB algorithm, Image-Based Fiber Orientation Calculator by university of Minnesota-Academic Use License was obtained [32], and images were cropped to have square dimensions in pixels with elimination to scale bars to have more accurate results. The MATLAB code processes the image to generate the orientation matrix [33] defined as:

Electric Field and Orientation Analysis
Nanofibers are formed between the 2-bars due to the high electric field. We introduced a finite element simulation for the proposed collectors during the study, showing both electric and electric field vectors between the 2 bars. The finite element package COMSOL Multiphysics®Version 5 (COMSOL Inc, Stockholm, Sweden) was used to demonstrate the fields profiles. The simulations were investigated using 3D module. SEM images were analyzed using Fourier Transform through the usage of MATLAB algorithm, Image-Based Fiber Orientation Calculator by university of Minnesota-Academic Use License was obtained [32], and images were cropped to have square dimensions in pixels with elimination to scale bars to have more accurate results. The MATLAB code processes the image to generate the orientation matrix [33] defined as: where Ii is the length of the fiber, Itot is the total sum of fiber lengths; θi is the angle between each fiber and the x-axis. The alignment scheme is determined through a factor called anisotropy index (α), where α = 0 for complete randomness of the nanofibers mats and α = 1 for completely aligned nanofibers. Anisotropy index is calculated from an equation relates the eigenvalues (λ1, λ2 and λ1 ≤ λ2), of the orientation matrix defined as follows: (2)

Characterizations Procedures
The morphology of PVDF NFs was observed by a Scanning Electron Microscope (JEOL JSM-6010LV-SEM). The diameter of NFs was analyzed using (Image J) software. The crystal phase of NFs was measured with X-ray Diffractometer (XRD) (Shimadzu Xlab 6100, Kyoto, Japan), and Fourier Transform Infra Red Spectrometer (FT-IR) (Vertex 70 FT-IR, Bruker, Billerica, MA, USA) was used for β phase content calculation. Simulation for alignment setup with orientation of electric field directions for plate and bars collectors was introduced using COMSOL finite-eliminate software. Degree of NFs alignment and anisotropy was observed using image-based fiber orientation and alignment calculator software program which is based on Fourier transform methods. Piezoelectric properties of the nanofiber mats are measured through a simple setup. The nanofiber mats of dimensions 2 cm × 2 cm were placed between two foil sheets, and pressed by different applied weights. Then, the generated voltage was detected through two connected shielded wires, pasted on the foil sheets, to a high impedance oscilloscope; Tektronix 3012 (Beaverton, OR, USA).

Fields Distribution Analysis
To show the electric field profiles, COMSOL Multiphysics was used to analyze the two collectors, to show the field profile on the normal collector and the 2 bars collector. Figure 2 shows the electric field distributions for the two types of collectors.
where Ii is the length of the fiber, Itot is the total sum of fiber lengths; θi is the angle between each fiber and the x-axis. The alignment scheme is determined through a factor called anisotropy index (α), where α = 0 for complete randomness of the nanofibers mats and α = 1 for completely aligned nanofibers. Anisotropy index is calculated from an equation relates the eigenvalues (λ1, λ2 and λ1 ≤ λ2), of the orientation matrix defined as follows:

Characterizations Procedures
The morphology of PVDF NFs was observed by a Scanning Electron Microscope (JEOL JSM-6010LV-SEM). The diameter of NFs was analyzed using (Image J) software. The crystal phase of NFs was measured with X-ray Diffractometer (XRD) (Shimadzu Xlab 6100, Kyoto, Japan), and Fourier Transform Infra Red Spectrometer (FT-IR) (Vertex 70 FT-IR, Bruker, Billerica, MA, USA) was used for β phase content calculation. Simulation for alignment setup with orientation of electric field directions for plate and bars collectors was introduced using COMSOL finite-eliminate software. Degree of NFs alignment and anisotropy was observed using image-based fiber orientation and alignment calculator software program which is based on Fourier transform methods. Piezoelectric properties of the nanofiber mats are measured through a simple setup. The nanofiber mats of dimensions 2 cm × 2 cm were placed between two foil sheets, and pressed by different applied weights. Then, the generated voltage was detected through two connected shielded wires, pasted on the foil sheets, to a high impedance oscilloscope; Tektronix 3012 (Beaverton, OR, USA).

Fields Distribution Analysis
To show the electric field profiles, COMSOL Multiphysics was used to analyze the two collectors, to show the field profile on the normal collector and the 2 bars collector. Figure 2 shows the electric field distributions for the two types of collectors. It can be observed that, in the conventional collector case, the electric field profile and vectors are not oriented in a certain direction; however, in the two-bar collector, the electric field vectors are It can be observed that, in the conventional collector case, the electric field profile and vectors are not oriented in a certain direction; however, in the two-bar collector, the electric field vectors are forming stretching forces in the gap between the two-bar. This can explain the differences in the fibers' common oriented direction. Surface morphology is shown in Figure 3.
forming stretching forces in the gap between the two-bar. This can explain the differences in the fibers' common oriented direction. Surface morphology is shown in Figure 3.

Orientation Analysis
Degree of alignment for the nanofibers mats (anisotropy index calculations) to prove the two-bar collector effect on the PVDF nanofibers alignment. The orientation matrix for the two-bar collector and conventional collector was obtained. Figure 4 shows the Orientation analysis for PVDF nanofibers collected on the 2 bars collector and the conventional collector. From results in Table 1, we notice an enhancement in the anisotropy index, i.e., an 82.45% improvement compared to the conventional collector.

Orientation Analysis
Degree of alignment for the nanofibers mats (anisotropy index calculations) to prove the two-bar collector effect on the PVDF nanofibers alignment. The orientation matrix for the two-bar collector and conventional collector was obtained. Figure 4 shows the Orientation analysis for PVDF nanofibers collected on the 2 bars collector and the conventional collector. From results in Table 1, we notice an enhancement in the anisotropy index, i.e., an 82.45% improvement compared to the conventional collector. forming stretching forces in the gap between the two-bar. This can explain the differences in the fibers' common oriented direction. Surface morphology is shown in Figure 3.

Orientation Analysis
Degree of alignment for the nanofibers mats (anisotropy index calculations) to prove the two-bar collector effect on the PVDF nanofibers alignment. The orientation matrix for the two-bar collector and conventional collector was obtained. Figure 4 shows the Orientation analysis for PVDF nanofibers collected on the 2 bars collector and the conventional collector. From results in Table 1, we notice an enhancement in the anisotropy index, i.e., an 82.45% improvement compared to the conventional collector.   Degree of alignment for the PVDF/Carbon nanotubes (CNTs) is calculated for two samples with 0.1% and 0.3% CNTs. The anisotropy index still shows an improvement in the two-bars setup for PVDF/CNTs samples when compared to conventional collector; however, the alignment percentage decreased due to the presence of CNTs. This may be explained by the effect of CNTs charge on the orientation of electrospun nanofiber. Figure 5 shows the Orientation analysis for PVDF/CNTs nanofibers collected on the conventional collector, while Figure 6 shows the Orientation analysis for PVDF/CNTs nanofibers collected on the two-bar collector. Alignment parameters are shown in Tables 2 and 3 for the conventional collector and the two-bar collector respectively. Figure 7 shows the complete comparison between the anisotropy index for all cases investigated through the study, it's noticeable the 2 bars collector effect on the degree of alignment for different cases.  Degree of alignment for the PVDF/Carbon nanotubes (CNTs) is calculated for two samples with 0.1% and 0.3% CNTs. The anisotropy index still shows an improvement in the two-bars setup for PVDF/CNTs samples when compared to conventional collector; however, the alignment percentage decreased due to the presence of CNTs. This may be explained by the effect of CNTs charge on the orientation of electrospun nanofiber. Figure 5 shows the Orientation analysis for PVDF/CNTs nanofibers collected on the conventional collector, while Figure 6 shows the Orientation analysis for PVDF/CNTs nanofibers collected on the two-bar collector. Alignment parameters are shown in Tables 2 and 3 for the conventional collector and the two-bar collector respectively. Figure 7 shows the complete comparison between the anisotropy index for all cases investigated through the study, it's noticeable the 2 bars collector effect on the degree of alignment for different cases.

Physical Analysis
XRD and FT-IR spectroscopy measurements were performed to analyze the crystal structure and content of the β-phase formation for the produced PVDF Nanofibers. The intensity of the main XRD peak corresponding to the PVDF β-phase (2θ = 20.6°) increased for the aligned nanofibers with twobars as compared to random nanofiber, whereas the non-polar α-phase (2θ = 40°) exhibited the opposite trend (Figure 8). The observed high content of the β-phase of the aligned nanofiber can be attributed to the stretching effect due to the electric field concentration between the two bars; when

Physical Analysis
XRD and FT-IR spectroscopy measurements were performed to analyze the crystal structure and content of the β-phase formation for the produced PVDF Nanofibers. The intensity of the main XRD peak corresponding to the PVDF β-phase (2θ = 20.6°) increased for the aligned nanofibers with twobars as compared to random nanofiber, whereas the non-polar α-phase (2θ = 40°) exhibited the opposite trend (Figure 8). The observed high content of the β-phase of the aligned nanofiber can be attributed to the stretching effect due to the electric field concentration between the two bars; when

Physical Analysis
XRD and FT-IR spectroscopy measurements were performed to analyze the crystal structure and content of the β-phase formation for the produced PVDF Nanofibers. The intensity of the main XRD peak corresponding to the PVDF β-phase (2θ = 20.6 • ) increased for the aligned nanofibers with two-bars as compared to random nanofiber, whereas the non-polar α-phase (2θ = 40 • ) exhibited the opposite trend (Figure 8). The observed high content of the β-phase of the aligned nanofiber can be attributed to the stretching effect due to the electric field concentration between the two bars; when nanofibers are oriented in the same direction and closely stacked, the polarization direction of the β-phase is also aligned, and thus, the β-phase content is self-reinforced (Figure 3b).
The IR vibrational bands observed at 612 cm −1 (CF 2 bending), 766 cm −1 (skeletal bending), and 795 cm −1 (CH 2 rocking) were due to the PVDF α-phase, while the vibrational band at 840 cm −1 , 879 cm −1 and 1270 cm −1 corresponded to the PVDF β-phase (Figure 9). According to Beer-Lambert law for the obtained IR spectra, the PVDF β fraction can be calculated by using the following equation: where F(β), represents the β phase content, and Aα and Aβ are the absorbance at 766 and 840 cm −1 respectively. By calculating the previous equation according to the obtained IR result, the β-phase content for the aligned NF was 86%, whereas the random NFs exhibited β-phase contents of 74.2%. nanofibers are oriented in the same direction and closely stacked, the polarization direction of the βphase is also aligned, and thus, the β-phase content is self-reinforced (Figure 3b). The IR vibrational bands observed at 612 cm −1 (CF2 bending), 766 cm −1 (skeletal bending), and 795 cm −1 (CH2 rocking) were due to the PVDF α-phase, while the vibrational band at 840 cm −1 ,879 cm −1 and 1270 cm −1 corresponded to the PVDF β-phase (Figure 9). According to Beer-Lambert law for the obtained IR spectra, the PVDF β fraction can be calculated by using the following equation: where F(β), represents the β phase content, and Aα and Aβ are the absorbance at 766 and 840 cm −1 respectively. By calculating the previous equation according to the obtained IR result, the β-phase content for the aligned NF was 86%, whereas the random NFs exhibited β-phase contents of 74.2%.

Piezoelectric Characterization
Regarding the piezoelectric analysis, each mat was exposed to pressures from different masses, and the generated peak-to-peak voltages were recorded. Both Figures 10 and 11 show the piezoelectric average peak-to-peak voltage of the generated nanofibers mats in the cases of nonaligned/aligned PVDFs with added CNTs at different applied weights. In the case of non-aligned nanofibers are oriented in the same direction and closely stacked, the polarization direction of the βphase is also aligned, and thus, the β-phase content is self-reinforced (Figure 3b). The IR vibrational bands observed at 612 cm −1 (CF2 bending), 766 cm −1 (skeletal bending), and 795 cm −1 (CH2 rocking) were due to the PVDF α-phase, while the vibrational band at 840 cm −1 ,879 cm −1 and 1270 cm −1 corresponded to the PVDF β-phase (Figure 9). According to Beer-Lambert law for the obtained IR spectra, the PVDF β fraction can be calculated by using the following equation: where F(β), represents the β phase content, and Aα and Aβ are the absorbance at 766 and 840 cm −1 respectively. By calculating the previous equation according to the obtained IR result, the β-phase content for the aligned NF was 86%, whereas the random NFs exhibited β-phase contents of 74.2%.

Piezoelectric Characterization
Regarding the piezoelectric analysis, each mat was exposed to pressures from different masses, and the generated peak-to-peak voltages were recorded. Both Figures 10 and 11 show the piezoelectric average peak-to-peak voltage of the generated nanofibers mats in the cases of non-

Piezoelectric Characterization
Regarding the piezoelectric analysis, each mat was exposed to pressures from different masses, and the generated peak-to-peak voltages were recorded. Both Figures 10 and 11 show the piezoelectric average peak-to-peak voltage of the generated nanofibers mats in the cases of non-aligned/aligned PVDFs with added CNTs at different applied weights. In the case of non-aligned nanofibers, as expressed in Figure 10, an enhancement of piezoelectric sensitivity of PVDF nanofibers mat due to the addition of CNTs can be observed. The sensitivity of the non-aligned nanofibers mat with no added CNTs was found to be 7.2 mV/g, and increased up to 8.8 mV/g at added 0.3 wt % of CNTs. The behavior of voltage-weight relation is mostly linear in the case of no CNTs, but starts to have some non-linearity in the case of 0.3 wt % CNTs at relatively higher applied masses. In the case of aligned nanofibers, as shown in Figure 11, the alignment shows an improvement in the sensitivity of piezoelectric response, compared to the same non-aligned fibers. In the case of increasing CNT concentrations, the linearity region becomes smaller at applied masses with higher sensitivity; up to 73.8 mV/g in case of 0.3 wt % of added CNTs, but at a lower range of applied weights down, i.e., to 100 g. However, the piezoelectric response starts to be saturated at higher range of applied masses. This can be helpful in using the nanofiber mat as an electronic switch when applying higher values of masses i.e., above 150 g. To check the impact of alignment only with no added CNTs, as shown in the no added CNTs curves in both Figures 10 and 11, it can be noted that the alignment leads to a better linear piezoresponse behavior due to the expected better alignment of polarized dipoles inside PVDF. nanofibers, as expressed in Figure 10, an enhancement of piezoelectric sensitivity of PVDF nanofibers mat due to the addition of CNTs can be observed. The sensitivity of the non-aligned nanofibers mat with no added CNTs was found to be 7.2 mV/g, and increased up to 8.8 mV/g at added 0.3 wt % of CNTs. The behavior of voltage-weight relation is mostly linear in the case of no CNTs, but starts to have some non-linearity in the case of 0.3 wt % CNTs at relatively higher applied masses. In the case of aligned nanofibers, as shown in Figure 11, the alignment shows an improvement in the sensitivity of piezoelectric response, compared to the same non-aligned fibers. In the case of increasing CNT concentrations, the linearity region becomes smaller at applied masses with higher sensitivity; up to 73.8 mV/g in case of 0.3 wt % of added CNTs, but at a lower range of applied weights down, i.e., to 100 g. However, the piezoelectric response starts to be saturated at higher range of applied masses. This can be helpful in using the nanofiber mat as an electronic switch when applying higher values of masses i.e., above 150 g. To check the impact of alignment only with no added CNTs, as shown in the no added CNTs curves in both Figures 10 and 11, it can be noted that the alignment leads to a better linear piezoresponse behavior due to the expected better alignment of polarized dipoles inside PVDF.

Conclusions
The work demonstrates the ability to generate highly-aligned PVDF nanofibers through staticalignment of the collector side of electrospinning setup. The synthesized PVDF nanofibers are studied with and without added in-situ CNTs. The two-bar collector system shows a great enhancement in the alignment degree for the PVDF nanofibers. Our results show that both alignment and added CNTs nanofibers, as expressed in Figure 10, an enhancement of piezoelectric sensitivity of PVDF nanofibers mat due to the addition of CNTs can be observed. The sensitivity of the non-aligned nanofibers mat with no added CNTs was found to be 7.2 mV/g, and increased up to 8.8 mV/g at added 0.3 wt % of CNTs. The behavior of voltage-weight relation is mostly linear in the case of no CNTs, but starts to have some non-linearity in the case of 0.3 wt % CNTs at relatively higher applied masses. In the case of aligned nanofibers, as shown in Figure 11, the alignment shows an improvement in the sensitivity of piezoelectric response, compared to the same non-aligned fibers. In the case of increasing CNT concentrations, the linearity region becomes smaller at applied masses with higher sensitivity; up to 73.8 mV/g in case of 0.3 wt % of added CNTs, but at a lower range of applied weights down, i.e., to 100 g. However, the piezoelectric response starts to be saturated at higher range of applied masses. This can be helpful in using the nanofiber mat as an electronic switch when applying higher values of masses i.e., above 150 g. To check the impact of alignment only with no added CNTs, as shown in the no added CNTs curves in both Figures 10 and 11, it can be noted that the alignment leads to a better linear piezoresponse behavior due to the expected better alignment of polarized dipoles inside PVDF.

Conclusions
The work demonstrates the ability to generate highly-aligned PVDF nanofibers through staticalignment of the collector side of electrospinning setup. The synthesized PVDF nanofibers are studied with and without added in-situ CNTs. The two-bar collector system shows a great enhancement in the alignment degree for the PVDF nanofibers. Our results show that both alignment and added CNTs

Conclusions
The work demonstrates the ability to generate highly-aligned PVDF nanofibers through static-alignment of the collector side of electrospinning setup. The synthesized PVDF nanofibers are studied with and without added in-situ CNTs. The two-bar collector system shows a great enhancement in the alignment degree for the PVDF nanofibers. Our results show that both alignment and added CNTs increased the formation of beta sheets inside the nanofibers, which correlates to a better polarizability inside the material. Consequently, the piezoelectric sensitivity of alignment PVDF doped with 0.3 wt % of CNTs was improved up to the range of 73.8 mV/g when applying weights up to 100 g, where the generated nanofibers mat can be used as a piezoelectric sensor. At higher applied weights, our synthesized nanofibers show saturation behavior which leads to the possibility of using our nanofibers as an electronic sensor with a steady output of generated voltage.