SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. EXPERIMENTAL
  5. RESULTS AND DISCUSSION
  6. CONCLUSION
  7. REFERENCES

Poly(vinyl) butyral (PVB) nanofibers (NFs) and carbon nanotube (CNT) reinforced PVB NF composites were developed by using the Forcespinning® technology. PVB was dissolved in a mixture of ethanol and methanol (7:3 wt/wt) at various concentrations, and the solutions were spun at rotational speeds varying between 3,000 and 9,000 rpm. The CNT/PVB solutions were prepared using the same solvent ratio with varying the concentration of CNTs. The results show that the diameter of the PVB fibers increased with increasing rotational speed; however the standard deviation of the fiber diameter distribution decreased. The morphology and thermal properties of the developed fiber systems were studied by DSC, TGA, Raman, and FTIR. The effect of CNT on the mechanical properties of the developed fibers was investigated by carrying out tensile tests at different strain rates. Raman and FTIR analyses indicate a noncovalent π–π stacking interactions and hydrogen bonding between CNT and the PVB NFs. Adding CNT to the PVB NF matrix resulted in improved tensile strength by 150%. POLYM. ENG. SCI., 55:81–87, 2015. © 2014 Society of Plastics Engineers


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. EXPERIMENTAL
  5. RESULTS AND DISCUSSION
  6. CONCLUSION
  7. REFERENCES

There has been a considerable amount of work in the literature on the use of polymer nanofibers (NFs) and NF composites in various applications such as, electronics, automobile, energy storage, military, aerospace, membrane science, filtration, drug delivery systems, tissue engineering, and biosensors to mention some [1-10]. Several processing methods exist for the production of NFs such as electrospinning, hydrothermal synthesis, template synthesis, phase separation, self-assembly, shear spinning, and other spinning methods (wet, dry, and melt spinning) [11]. Electrospinning could be considered a simple and low cost process given that only a syringe and a high voltage source are needed for lab scale production, though the process itself presents many limitations such as the need of high electric fields and typically the need of a dielectric solution besides the intrinsic suitable vapor pressure, viscosity, and surface tension to promote optimum fiber formation. These limitations render electrospinning as a complex and costly method for industrial potential. In general, electrospinning has been a widely used method to produce NFs at a lab scale. In the case of melt-blown process, which is a traditional melt spinning process, micro-fibers can be easily produced but in order to produce NFs, higher speeds of heated air are necessary which makes this process impracticable especially for industrial production. Forcespinning® (FS) technology (developed by Lozano et al.), as well as similar centrifugal spinning methods have the opportunity to offer a high production rate of NFs with low cost and a broader choice of materials given the absence of electric fields [11-21]. Several systems have been recently produced either from solution or melt, such as polyamide, polyethylene oxide, polylactic acid, polycaprolactone [15], polypropylene [16], ultra-high molecular weight polyethylene, (2,5-bis(2′-ethyl-hexyl)−1,4-phenylenevinylene) [12], poly(3-hexylthiophene (P3HT), Teflon®AF [17], cellulose, poly(vinylidene fluoride) [14], and indium-tin oxide [18] to mention some. All of the above have produced average fiber diameters in the nano or submicron range.

In this work, we use the FS method to prepare poly(vinyl) butyral (PVB) NFs and carbon nanotube (CNT)/PVB NF composites for potential applications such as electronic technology that require the use of polymer/ceramic nanocomposites with enhanced thermal conductivity, low coefficient of thermal expansion, and improved dielectric breakdown [22]. PVB is a semi-crystalline polyester with low cost, good recyclability, clarity, toughness, flexibility, and excellent adhesion properties [23-25]. In addition, PVB is a good solute in alcohols and aromatic solvents which makes PVB solutions nontoxic and desirable for use in industry. It is mainly used as adhesive (binder) and matrix in ceramic based composites due to its compatibility and processability [26-28]. The high adhesion properties of PVB make it a good candidate for use in nanocomposites where the adhesion between the matrix and filler is crucial in determining the properties and performance of the nanocomposites [26-29]. For example, PVB was used as a matrix for boron nitride nanotubes (BNNT) to fabricate composites for electronic applications. The results showed a significant improvement in the thermal conductivity and dielectric properties of the composites. The dielectrical breakdown of the PVB/BNNT was also improved [22]. Although an extensive work has been reported on the processing and characterization of PVB, a limited number of studies are available on the production of PVB NFs [23, 24]. PVB has been used as a polymer matrix for CNT to fabricate nanocomposites with enhanced electrical and thermal properties [30-32]. CNT reinforced polymer composites had lately been a source of intense research given their attractive properties such as high aspect ratio of CNT, good electrical conductivity, good mechanical strength, and high thermal stability [33-38]. CNT reinforced PVB composites have also been produced to enhance hardness, coefficient of friction, and scratch resistance of PVB [29].

A recent study utilized a surface-draw method to produce CNT reinforced PVB fiber composites, where the CNT content was varied from 0 to 80 wt%. The diameter of the developed fibers varied from 10 to 100 µm. Compared to pure PVB fibers, it was found that the CNT/PVB composites with 7.4 wt% of CNT concentration showed 127% increase in the tensile strength when compared to pure PVB fibers [32]. However, the diameter of the composite fibers is higher than 10 µm. In this work, PVB and CNT/PVB NF composites are fabricated by using the FS method and the process variable effects on the mechanical and electrical properties of PVB NF matrix were systematically investigated. The main thrust of the article is the large-scale approach for the production of high tensile strength CNT reinforced PVB NFs.

EXPERIMENTAL

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. EXPERIMENTAL
  5. RESULTS AND DISCUSSION
  6. CONCLUSION
  7. REFERENCES

Materials and Methods

PVB (average Mw 70,000), anhydrous ethanol, methanol, sulfuric acid (H2SO4, 98%), anhydrous potassium bromide (KBr > 99%), and nitric acid (HNO3, 70%) were purchased from Sigma Aldrich. Multi-walled carbon nanotubes (outer diameter 30–50 nm) were purchased from Cheap Tubes Inc. Deionized water (18 MΩ cm) was produced using Mill-Q (Millipore Ltd., UK). The filters used for functionalized CNT filtration are 0.45 µm filter paper from Millipore. A bath sonicator model cole-parmer 8890 was used. For the spinning of the fiber, a lab scale Cyclone™ L-1000M from FibeRio Technology Corp. was utilized.

CNTs were functionalized to optimally disperse CNTs in the alcohol-based solvent and to increase the affinity of CNT to the polymer matrix. CNTs are often bundled together due to strong van der Waals interactions between the nanotubes [39]. Here, CNTs were functionalized with 3:1 (vol/vol) H2SO4 and HNO3 at 70°C to generate carboxylic acid groups according to Lozano et al. [39]. After functionalization, the CNTs were well dispersed in water, and alcohols without precipitation for several months. About 500 mg of pristine CNTs were added to 150 mL of a mixture of concentrated sulfuric acid and nitric acid with a volume ratio of 3:1 followed by sonication at 50°C for 4 h and refluxation at 70°C for 12 h. The mixture was then diluted by a large amount of deionized water and filtrated, washed with deionized water until the filtrate reached pH 7, and then dried at 120°C for 24 hr.

Fiber Spinning

FS was conducted utilizing a lab scale system, the Cyclone L-Tool (Manufactured by FibeRio Technology, Corp.) with an angular velocity capability of 1,000 to 20,000 rpm. Figure 1a illustrates a schematic of the FS setup, and Fig. 1b shows the photograph of the spun PVB fibers and CNT/PVB (insert) nonwoven fiber mats (spinning duration 10 sec) obtained by using our lab scale system, in which the fiber web is collected on the designed collector system.

image

Figure 1. (a) Schematic of Forcespinning® setup. (b) Photograph of spun PVB NF web (cycle time 10 sec, arrow labeled). (b) Inserted images: (b1) Pure PVB, (b2) 3 wt% CNT in PVB, (b3) 6 wt% CNT in PVB, (b4) 9 wt% CNT in PVB.

Download figure to PowerPoint

In the FS technique, centrifugal force is used to extrude polymer solutions or melts. Fiber jets are formed by high rotational speeds of a spinneret. The rotation of the spinneret at high speeds (up to 20, 000 rpm) drives the fluid through the orifices. When the centrifugal force and associated hydrostatic pressure exceeds capillary forces that tend to restrict the flow of fluid in the orifice, a jet of polymer solution is ejected. Reduction of fiber diameter occurs by the inertial drag between the fiber and the atmosphere as the polymer solution jet dries. The polymer mass flow rate through the spinneret is partially governed by a pressure driven flow from the outward centrifugal force acting on the solution at the spinneret entrance. Though the phenomenon of centrifugal force-based spinning is based on the principal of cotton candy machine, there are few reports demonstrating its capabilities as a NF manufacturing technology [11-18].

PVB and CNT/PVB NF Composite Preparation

PVB was dissolved at various concentrations in 7:3 (ethanol: methanol, wt/wt) alcohol mixture, and spun at various rotational speeds to determine the optimal range for low fiber diameter. For CNT/PVB composites, PVB was added to the suspension of CNT followed by stirring the mixture for 24 hr under inert atmosphere. About 2 mL of the solution were injected into the spinneret prior to FS. The needle-based spinneret equipped with 30 gauge half-inch regular bevel needles (Beckett-Dickerson) was used. The rotational speeds were varied from 3,000 to 9,000 rpm. A deep dish fiber collector with equally distanced vertical steel pillars was used to collect fibers. After collection, the fibers were covered and stored under desiccation.

Nanofiber Characterizations

The morphology of NFs was investigated using scanning electron microscopy (SEM) and transmission scanning electron microscopy (STEM) with a Sigma VP Carl Zeiss, Germany. For STEM characterization, samples of CNT/PVB NFs were embedded in epoxy resin (LR White-Medium Grade, Ladd Research), and microtomed at room temperature into sections of 60 nm thick. The samples were then examined using STEM at 27 kV. Thermogravimetric analysis (TGA), differential scanning calorimetry (DSC), and dynamic mechanical analysis (DMA) were performed using TA-Q series equipments, TGAQ500, DSCQ100, and DMAQ800, respectively. About 10 mg samples were used in the TGA and DSC experiments which were performed at a heating rate of of 10 °C min−1. DSC results were obtained after heating and cooling at the same rate. Raman spectrum was performed using Bruker Senterra Raman spectroscopy with a 785 nm excitation laser. Fourier transform infrared spectroscopy (FTIR) was performed using a Bruker IFS 55 Equinox FTIR spectrophotometer with KBr pellets. For DMA testing, the samples were cut with a razor into rectangular strips of 2 × 1 × 0.5 cm (length × width × thickness) mats and gripped using a film tension clamp with a clamp compliance of about 0.2 µm N−1. All tensile tests were conducted in controlled strain rate mode with a preload of 0.01 N and a ramp rate of 0.01 N min−1. To reduce variability, the mats were about the same weight and same GSM (grams per square meter). The DMA plots were selected from curves that overlapped at least twice for each sample.

RESULTS AND DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. EXPERIMENTAL
  5. RESULTS AND DISCUSSION
  6. CONCLUSION
  7. REFERENCES

The fiber morphology and diameter are influenced by several experimental variables including orifice diameter spinneret to collector distance, viscosity of the solution, and rotational speed (rpm). Previous work has shown that there is a minor effect of the needle-collector distance on fiber diameter compared to the concentrations and the rotating speeds [11-18].

The results show that the polymer concentration has a stronger effect on fiber characteristics than other process. Beads were formed at concentrations lower than 12 wt%, which is a result of the interplay between three factors, surface tension, the applied force, and the viscoelastic forces. Surface tension tends to minimize the surface area of the polymer jet exiting the orifice leading to bead formation. The centrifugal forces and viscoelastic forces favor the fiber formation. The increase in polymer concentration (or viscosity) leads to fiber formation with larger diameters. For example, increasing the concentration from 11% (Fig. 2a) to 12% (Fig. 2b), to 13% (Fig. 2c), and to 14% (Fig. 2d) led to a significant increase in the average diameters from 400 nm to 550 nm, to 720 nm and to 1 µm (Fig. 2f), respectively. The diameter distributions of these NFs are shown in Fig. 2e. The fiber diameter distributions appear to follow a log-normal function; 12% PVB sample exhibits an average log(diameter, micrometers) of −0.21 (617 nm) with a standard deviation of 0.1. To ensure reproducibility of the results, two batches of fibers were made independently for each condition, and the SEM images were analyzed. The average fiber diameter and other data related to fiber dimensions are based on the average of all measurements that were taken (more than a few hundred per sample with random image sampling). The change in concentration can influence the fiber formation in different ways. In fact, dynamic shear experiments revealed nearly an order of magnitude reduction in shear viscosity of PVB from 1 Pa sec at 15% to 0.1 Pa sec at 11% at 50 Hz of shear rate. So the thinner fibers obtained at lower polymer concentration could be due to lower viscosity and/or additional stretching before sufficient evaporation (higher solvent content) takes place for the onset of NF formation. However, the formation of beads at a concentration lower than 12% could not be avoided. Increasing the temperature could lead to a faster solvent evaporation but it will further decrease the viscosity.

image

Figure 2. SEM images of spun PVB NFs with PVB concentrations of 11wt% (a), 12 wt% (b), 13 wt% (c), and 14 wt% (d). (e) Diameter distributions of PVB NFs at various PVB concentrations with spinning speed of 7,000 RPM. (f) Shows the dependence of fiber diameter with the concentration of PVB.

Download figure to PowerPoint

The rotational speed of the spinneret also had an important effect on fiber formation [11-18]. Figure 3a–d shows the SEM images of PVB fibers prepared by FS at different rotational speeds of 3,000, 6,000, 7,000, and 9,000 rpm. At a speed of 3,000 rpm (Fig. 3a), the fibers were 528 nm in diameter. Increasing the speed to 9,000 rpm resulted in increasing the yield of NFs, while the average fiber diameter was slightly increased to 606 nm (Fig. 3e). Higher spinneret speeds can produce fibers with a narrower diameter distribution but much thicker fibers due to the increase in the Rossby number [13].

image

Figure 3. Spun 12 wt% PVB solution at 3 K (a), 5 K (b), 7 K (c), and 9 K (d) RPMs. (e) Diameter distributions of spun 12 wt% PVB solution at various rotating speeds.

Download figure to PowerPoint

The spun CNT/PVB NFs were gray or black, depending on the content of the CNT (Fig. 1b inserted images). The concentration of PVB is 12 wt%, and the rotational speed is 7,000 RPM. Figure 4a–d shows SEM images of CNT/PVB NFs at different CNT concentrations (3, 6, 9, 10 wt%). It can be noted that the surface morphology of the composite NFs with concentrations higher than 9 wt% of CNT is rougher than those of pure PVB NFs (Fig. 3). In the STEM image of the 9 wt% CNT sample (Fig. 4d), one can observe that CNT are extruded out from the polymer matrix, and fully wetted by the polymer matrix. The CNTs are well dispersed in the polymer matrix without agglomeration. Figure 4e displays the diameter distribution of the spun CNT/PVB fibers, as indicated the fibers have 600–900 nm in average diameter.

image

Figure 4. SEM images of CNT/PVB NF composites, (a) 3 wt% CNT, (b) 6 wt% CNT, and (c) 9 wt% CNT. (d) STEM of (c), CNTs are labeled with black arrows. (e) Diameter distribution for spun CNT/PVB NF composites.

Download figure to PowerPoint

Thermophysical Characterization

Figure 5 shows TGA and DSC curves of PVB NFs and CNT/PVB NF composites. The experiments were carried out over a temperature range of (500–600°C) and at a heating rate of 10 °C min−1. The DSC plot of PVB NFs shows a degradation peak at 360°C, whereas the CNT/PVB composite shows degradation peak at 373°C, which is about 10°C higher than the pristine PVB fibers indicating an interaction between CNT with PVB. The TGA results show that PVB NFs are fully degraded below 450°C, while the CNT is stable in the nitrogen atmosphere. Based on the weight loss observed in the TGA curves, the amount of CNT loaded in the NFs were calculated to be 6% and 9%. The TGA results indicate interaction between CNTs and the PVB matrix.

image

Figure 5. TGA (a) and DSC (c) curves of PVB NFs and CNT/PVB NF composites at a ramping rate of 10 °C min−1. (b) The derivation weight loss curves of TGA data from (a).

Download figure to PowerPoint

Raman Spectroscopy Analysis

A typical Raman spectrum (Fig. 6a) of CNT indicates two characteristic peaks. The G-line peak at 1580 cm−1 which corresponds to the high-frequency Raman-active E2g mode of graphite, and the D-line at 1350 cm−1 sophistically attributed to disorder-induced carbon features arising from finite particle size effect or lattice distortion [40]. The results show that for the CNT/PVB composite NFs, the peak positions of the CNT remain unchanged suggesting that the lattice structure of the CNT was not significantly affected by the presence of the polymer matrix. However, the intensity of G-line decreased in the presence of PVB indicating that the intrinsic electronic properties of the hexagonal carbon skeleton were substantially affected while the C[BOND]O[BOND]C vibrational peaks, positioned at 850 cm−1, were enhanced. Based on these observations, we can speculate that there are π–π stacking interactions between the circular ester carbon network of PVB and the hexagonal carbon network of CNT, resulting in the difference in Raman spectroscopy. Also, as shown in Fig. 6, the peaks of the PVB polymer NFs disappeared except for the C[BOND]O[BOND]C vibrational peak because of the lattice change in the PVB polymer chains after adding CNT. This confirms that there is a molecular interaction between the PVB and CNT where PVB interacts with the surface of the CNT by π–π interaction.

image

Figure 6. Raman (a) and FTIR (b) spectra of CNTs (1), PVB (2), 6 wt% CNT/PVB (3), and 9 wt% CNT/PVB (4) NFs. Dashed lines were added to the FTIR spectra for clarity. (c) FTIR spectra in the wavenumber ranges of 1500–1800 cm−1.

Download figure to PowerPoint

FTIR Analysis

Figure 6b compares the FTIR spectra for (a) carboxylic functionalized CNT, (b) pristine PVB, and CNT/PVB composites. The intrinsic peak of CNT is at 1631 cm−1, assigned to the C[DOUBLE BOND]C stretching vibration mode associated with sidewalls of the CNT. After the acid treatment, an additional peak was observed at 1713 cm−1, which was assigned to the carboxyl group stretches (C[DOUBLE BOND]O). The broad band at 3450 cm−1 was attributed to [BOND]OH stretching in carboxylic acid group. PVB NFs has also a peak at 1735 cm−1, assigned to the ester bond of the isomers of PVB, and the intensity of this peak is close to the1631 cm−1 peak. As indicated, after adding CNT to the PVB matrix, the peak intensities of the carboxyl group of the PVB and functionalized CNT decreased while the intensity of the C[DOUBLE BOND]C stretching vibration increased slightly. Based on these examinations, we can prove that the interactions between PVB and CNT are not just the π–π interaction, but also the interactions of the functional groups.

Mechanical Characterization

Tensile testing was performed to study the mechanical properties of the PVB NFs and the CNT reinforced fibers. Stress–strain curves of the CNT/PVB NF composites are shown in Fig. 7. It can be seen that the addition of CNT to PVB matrix does improve the tensile strength of the composite NF mat. The tensile strength of the 9 wt% CNT composite shows a peak value of 5 MPa at 20% strain which indicates a 150% improvement in tensile strength when compared to the PVB NF mat. As a result of the observed interactions between the matrix and the CNT, the tensile strength of the NF composites was expected to increase with increasing CNT content. The strain at break becomes smaller with the increase in CNT content in the PVB NF matrix. In fact, the strain at break decreases with increasing CNT content in the PVB NF composites. Results reported in the literature showed that improvement in tensile strength of most CNT/polymer composites are accompanied with a reduction in strain at break of the composite [41-43]. Even though, these results seem high (Fig. 7) when compared to traditional values reported on PVB microfibers (20 MPa) [30], the values are considered to be low as the potential provided by the CNT, this gives the loose overlapping of small diameter fibers and the isotropic orientation of the fibers within the mat.

image

Figure 7. Tensile strength curves of pure PVB (1), 3 wt% CNT/PVB (2), 6 wt% CNT/PVB (3), and 9 wt% CNT/PVB (4).

Download figure to PowerPoint

CONCLUSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. EXPERIMENTAL
  5. RESULTS AND DISCUSSION
  6. CONCLUSION
  7. REFERENCES

A large-scale production of PVB NFs and CNT reinforced PVB composite NFs was developed using the FS method. Raman and FTIR analyses suggest that there are interactions between CNT and the PVB NF matrix through π–π interactions and hydrogen bonding. The results show that PVB fibers were obtained with a diameter ranging from 400 nm to 1 µm. Increasing the PVB concentration resulted in increasing fiber diameter. If PVB concentration was reduced, fibers with lower diameters were obtained though the overall structure of the material contained large amount of beads. CNT/PVB fiber composites were also obtained at different CNT content and the results show that this new NF production method is suitable not only for producing polymer NFs but also for preparing CNT reinforced NFs at considerable larger concentrations than any of the other fabricating methods. Compared to pristine PVB NFs, the carbon nanotubes reinforced composite show a 150% improvement in tensile strength. The results show that FS is an attractive method to develop large-scale production of NF and in particular NF composites with high filler content.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. EXPERIMENTAL
  5. RESULTS AND DISCUSSION
  6. CONCLUSION
  7. REFERENCES