3.1. Material Production Method
As mentioned in the introduction, the production methods used for the fabrication of pure CNT-TPU composite fiber reported so far, were based on simple mixing of polymer melt/solution with the nanofiller powder which resulted in low nanofiller content and therefore low conductivity [22
]. So as to solve this issue, in the following paper, a different method was proposed. A weighed amount of CNTs was sonicated for two hours in excess amount of tetrahydrofuran (THF) (CNTs concentration was not higher than 2 wt%). A 2 wt% of surfactant in relation to nanocarbon mass, was also added to the solution, so as to facilitate deagglomeration and suspension of the CNTs. Subsequently, the weighed amount of TPU was mixed with the solution for five hours at 50 °C. Afterwards, the samples were left under the fume hood for 48 h to evaporate the solvent. Thus, the final content of carbon nanotubes in the composite was controlled only by the weight ratio of CNTs and polymer added to the solvent. Taking into account an extremely small amount of surfactant, this assumption should not introduce any significant error.
After evaporation of the solvent, the composite was palletized and formed into a fiber using a hot mixing extrusion process. So as to obtain comparable results, the nozzle of 1.5 mm diameter was used for the production of all composites. The regulated temperatures of the heating zones were set at 140 ℃ and 160 °C, for the first and second heating zone, respectively (Figure 1
). Lowering of the temperatures resulted in insufficient plasticization of the matrix material, which hindered the extrusion process and made the material inhomogeneous. On the other hand, too high temperatures resulted in thermal degradation of the polymeric matrix. The optimization of the fiber extrusion process enabled a continuous formation of the highly homogeneous fibers of any length and high diameter uniformity along the length.
3.2. Composite Characterisation
The composite production method proposed above enabled the formation of a composite fiber with a much higher content of carbon nanotubes than reported so far. The maximum weight percent of carbon nanotubes amounted to 40% as compared to 5% w
of CNT reported previously [22
]. Further, the increase in the content of nanotubes in the fiber was associated with a visible deterioration of elastic properties. At 50% w
of nanotubes, the fiber crumbled and it was not possible to extrude it continuously.
presents scanning electron microscope images performed on the cross-section of the sample with 40% w
of CNT in TPU. The images show high isotropy of the fiber structure at the microscale (Figure 2
a,b) as well as clear presence of carbon nanotubes (Figure 2
c) at the nanoscale. It is also visible that CNTs are very uniformly distributed in the polymer matrix.
The same material was also subjected to Raman spectroscopy which is an analytical tool widely used for the characterization of carbon nanotubes (Figure 3
). The characteristic features of CNT Raman spectra are: RBM (radial breathing mode), D-band, G-band, and 2D (or G’)-bands. RBM peaks appear at low wavenumbers and are visible only for a high concentration of single wall carbon nanotubes (SWCNT). D-band observed at approximately 1340 cm−1
is associated with the presence of disordered and amorphous carbon [27
]. G-band appears around ~1580 cm−1
and is related to in-plane carbon-carbon bond stretching [28
]. Last feature characteristic for CNT materials is an overtone to D-feature, known as 2D-band or G’-band. This is a peak observed for all sp2
bonded carbons, qualitatively not related to structure disorder [29
]. Finally, the disorder and impurity of the CNT materials are often assessed based on intensity ratio for and D and G peak ID
. The smaller it is the better the quality of the material.
The analysis of the Raman spectroscopy results presented in Figure 3
, shows that all features characteristic for CNT materials are present in the CNT-TPU spectrum, indicating a clear presence of these materials. It is quite interesting to find high intensity RBM peaks testifying a presence of SWCNTs in the industrial grade CNTs batch used for the manufacture of the fibers. Finally, it is worth mentioning that D-band is broad and ID
intensity ratio is high, which could indicate low graphitization and purity of the material. However, taking into account that all CNTs are coated with TPU it should be rather understood as a feature characteristic for the composite.
The as-produced nanotube-polymer fibers have been further subjected to the tests of electrical conductivity, current carrying capacity and stress-strain tensile tests. So as to understand the influence of the CNT content on the properties of the fibers. The results obtained for the composite composition containing 40% w/w of carbon nanotubes were also compared to the results obtained for fiber containing 20% w/w of CNT in TPU.
The electrical conductivity of the 40 wt% CNT sample amounted to 671 ± 22 S/m. This is over an order of magnitude better result than for previously reported CNT/TPU fibers. This is also a comparable result to other non-doped and non-annealed CNT-polymer composite fibers [12
]. Decrease in a CNTs content to 20 wt%, results in a drop of conductivity by over 2 orders of magnitude to 4.2 ± 0.7 S/m. However, this is still a very good result as compared to other CNT/TPU fibers.
presents the results of current carrying capacity tests, performed in a step mode i.e., DC current has been increased by 0.01 A every 1 s. It is visible that for the 40% w
content of nanotubes, the maximum current reached 1.25 A, while for 20% w
1.8 A. This result is quite unexpected taking into account the conductivities of the fibers. However, it is possible that the differences in the density of the materials and different heat removal conditions are responsible for such discrepancies.
Nevertheless, it should be rather noticed here that both fibers have shown very good results and failed at over 1 A. Taking into account poor thermal and electrical conductivity of TPU it is a very impressive outcome enabling many electronic applications.
Finally, stress-strain tests performed on both types of samples presented in Figure 5
demonstrate that the fibers show classical elastic and plastic deformation regions observed for pure CNT fibers and other CNT-polymer composite fibers [21
]. However, what is more important mechanical properties of the fibers may be easily controlled by changing the CNT content in the composite. Clearly, the 40 wt% composite shows higher strength and lower maximum elongation, while the decrease in strength is correlated with a higher elasticity of the 20 wt% fibers.
It is also worth noting that the maximum elongations of 35% for 40% w
CNT in TPU and 70% for 20% w
CNT in TPU are very high as for CNT materials [30
]. This property should be also particularly useful in textile applications.
Finally, the proposed composites are also very lightweight as the density of the composite fibers amounts to 1.1 ± 0.1 g/cc only. The above presented analysis shows that using the proposed method of CNT-TPU fiber manufacture it is possible to produce highly conductive light-weight composite materials with the potential to be applied in modern electronics including smart textiles.
Application of the composite in any electronics area requires separate extensive research. Taking into account that the produced composites may be particularly interesting for smart textiles applications we approached this area as an example. Firstly, 10-metres-long composite with 40% w
of CNTs has been manufactured and wound on a reel (Figure 6
a). Secondly, the issue of introduction of such composites into the fabrics, has been considered.
The revision of the literature indicates that there are many methods by which the conductive elements may be integrated into the fabrics. However, they may be generally divided into two main types which include weaving/sewing or depositing of the coating layer on the fabric [6
]. In the case of our CNT-TPU composites both methods are possible.
As shown in Figure 6
b the self-standing composite fibers were both sewn into a fabric and transferred onto fabric by a thermal process, where the polymer softens and attaches to the woven classical yarns. Such a “textile print” was performed using a simple iron as shown in Figure 6
c,d. The integrated fibers are very strongly attached to the fabric (cannot be removed by hand), while fatigue testing showed that a bending of the material 100 times by at least 170° did not cause any mechanical damage to the conductive pathway or change in electrical conductivity.