NANO LETTERS
Hot-Drawing of Single and Multiwall Carbon Nanotube Fibers for High Toughness and Alignment
2005 Vol. 5, No. 11 2212-2215
P. Miaudet,† S. Badaire,† M. Maugey,† A. Derre´,† V. Pichot,‡ P. Launois,‡ P. Poulin,*,† and C. Zakri*,† Centre de Recherche Paul Pascal - CNRS, AVenue Schweitzer 33600 Pessac, France, and Laboratoire de Physique des Solides, UniVersite´ Paris Sud, 91405 Orsay, France Received July 22, 2005; Revised Manuscript Received September 5, 2005
ABSTRACT We report a new hot-drawing process for treating wet-spun composite fibers made of single- and multiwall carbon nanotubes and poly(vinyl alcohol). As shown in previous reports, untreated composite nanotube fibers exhibit a very large strain-to-failure, and their toughness, which is the energy needed to break the fibers, exceeds that of any other known materials. However, untreated composite nanotube fibers absorb a very small amount of energy at low strain and become degraded in humid conditions. In this work, we use hot-drawing treatments, a concept inspired from textile technologies, to improve the properties of nanotube/PVA fibers. This treatment yields a crystallinity increase of the PVA and an unprecedented degree of alignment of the nanotubes. These structural modifications lead to a markedly improved energy absorption at low strain and make the fibers resistant to moisture. Hot-drawn nanotube/PVA fibers hold great potential for a number of applications such as bulletproof vests, protective textiles, helmets, and so forth.
Fibers and yarns are among the most promising forms for using nanotubes on a macroscopic scale mainly because, in analogy with high-performance polymer fibers, they allow nanotubes to be aligned and then weaved into textile structures or used as cables. A number of methods have been proposed to produce fibers composed entirely of nanotubes or a large fraction of nanotubes.1-7 The quickly increasing production of nanotubes is currently allowing the exploration of new treatments and processes that could lead to viable and useful nanotube textile technologies in the near future. Dalton et al. have recently reported high toughness of fibers made of single-wall nanotubes (SWNT) and PVA.8 These “super-tough” fibers could absorb 570 J/g for a strainto-failure of about 100%. This toughness is 1 order of magnitude larger than that of polyaramide fibers such as Kevlar or Twaron. In this work, using methods described already, we demonstrate that untreated SWNT fibers are capable of absorbing 870 J/g with a strain-to-failure up to 430%. In addition, we show for the first time that supertough fibers, with a toughness of 690 J/g and strain-to-failure of 340%, can also be spun out of multiwall nanotubes (MWNT), thereby providing more opportunities for the use of various materials. However, in several applications, such as bulletproof vests, absorption energy at lower strain is required. As another weakness, classical nanotube/PVA fibers * Corresponding authors. E-mail:
[email protected]; zakri@ crpp-bordeaux.cnrs.fr. † CNRS. ‡ Universite ´ Paris Sud. 10.1021/nl051419w CCC: $30.25 Published on Web 10/11/2005
© 2005 American Chemical Society
Figure 1. 10-m-long nanotube fiber collected on a small winder. The diameter of the fiber is 30 µm.
swell in water and lose strength in humid conditions. 9 Polyaramide fibers are less water-sensitive and absorb 35 J/g for a strain of only 3%.10 At such a strain, super-tough nanotube/PVA fibers have not yet absorbed 10 J/g. It is thus critical to design fibers that can absorb energy at lower strain and can be water resistant. We report that both properties can be achieved by hotdrawing SWNT as well as MWNT fibers. This treatment improves the nanotube and PVA alignment and increases the PVA crystallinity. The latter has been shown to be critical to enhancing stress transfer between nanotubes and PVA in composite materials.11 In addition, it prevents PVA from swelling in water. These so-called “hot-stretched” fibers
Figure 2. (a) Two-dimensional detector image X-ray data of a super-tough-type fiber. The broad ring indicates an alignment of the PVA chains of about (27° (b) Similar to a, for a hot-stretched nanotube fiber. The sharp spots reflect better crystallinity and better alignment of the PVA chains, which is only (4.3° here. Data analysis of such spectra is detailed in refs 12 and 13. (c) Stress vs strain curve of a super-tough SWNT fiber. The toughness is 870 J/g and the strain-to-failure reaches 430%. Inset: zoom on the data at low strain. (d) Similar to c for a hot-stretched fiber. The strength is 1.6 GPa; the toughness is 55 J/g. It is lower than that in c, but the absorption energy at low strain is significantly larger. The data shown are obtained with fibers spun with Elicarb nanotubes.
exhibit a mechanical behavior markedly different from that of any nanotube fibers reported previously, including thermally untreated super-tough fibers. Several hot-stretched nanotube fibers were tested; we found values of tensile strength between 1.4 and 1.8 GPa, strain-to-failure between 6 and 12%, and toughness between 40 and 60 J/g. We spin fibers via a continuous method with the fundamental mechanism described in the literature.2 We use nanotubes from three providers: Elicarb SWNT from Thomas Swan, HiPco SWNT from Carbon Nanotechnologies Inc., and MWNT from Arkema. Elicarb and HiPco SWNTs (0.35 wt %) are dispersed in water using sodium dodecyl sulfate (SDS) as the dispersant (1 wt %). MWNTs are dispersed at a higher concentration (0.9 wt %), also in an aqueous solution of SDS (1.2 wt %). The dispersions are homogenized by a sonication treatment. The fibers are spun continuously by injection of the homogeneous dispersions in the coflowing stream of an aqueous solution of PVA (5 wt %, MW 150 000). The spinning rate is about 6 m/min. A 10-m-long fiber collected onto a small winder is shown in Figure 1. The resultant Nano Lett., Vol. 5, No. 11, 2005
fibers contain an equal weight fraction of PVA and nanotubes. The fibers are dried and then drawn to 850% in a flow of hot air at 180 °C, a temperature well above the PVA glass transition. Their structure is characterized by X-ray diffraction (XRD). The degree of alignment of the polymer chains and nanotubes is deduced directly from Gaussian fits of the angular distribution of the scattered intensity at welldefined wave vectors as described in detail in refs 12 and 13. The mechanical properties of the fibers are characterized under tensile load using a Zwick Z2.5/TN1S instrument. After performing several tests, we found reproducible results within a 30% margin. Untreated SWNT and MWNT fibers exhibit strain-to-failure above 250% regularly. These untreated fibers exhibit a super toughness but a weak energy absorption at low strain. Elicarb and HiPco nanotubes lead to very similar results and cannot be distinguished in several tests. The best result was nevertheless obtained with an Elicarb nanotube fiber. As shown in Figure 2, its strain-to-failure reaches 430%, which is more than 4 times the highest strainto-failure found previously for a nanotube fiber.8 Conse2213
Figure 3. Stress vs strain curve of a super-tough MWNT fiber. The toughness is 690 J/g and the strain-to-failure reaches 340%.
quently, such a fiber exhibits an extremely high toughness of about 870 J/g, which is to our knowledge the highest toughness reported for a material. Untreated MWNT fibers compare well with SWNT fibers even though the best results are slightly lower. The maximal strain-to-failure we measured was about 340% with a toughness of about 690 J/g. The mechanical characterization of this fiber is shown in Figure 3. Nevertheless, all of these fibers exhibit a weak energy absorption at low strain. It is typically less than 10 J/g for strain up to 10%. In addition, they swell in water and lose their mechanical properties. After hot-drawing, their behavior is markedly different. The best modulus and tensile strain are obtained for hot-
stretched HiPco and Elicarb fibers: about 45 GPa for the modulus and 1.8 GPa for the tensile strength. The modulus and strength of hot-stretched MWNT Arkema fibers compare well with SWNT although they are slightly weaker. The modulus is about 35 GPa and the strength is about 1.4 GPa. XRD characterizations, shown in Figure 2, confirm the strong effect of hot-drawing. The characteristic data for two types of fibers are shown: a super-tough type fiber and a hotstretched fiber. As shown in Figure 4, the alignment of the PVA chains for an untreated super-tough fiber is about (27°. As observed in previous nanotube fibers, it matches the alignment of the nanotubes.12,13 Hot-stretched fibers exhibit a significant structural improvement. As shown in Figures 2 and 4, the PVA chain alignment is (4.3° and (6.3° in SWNT and MWNT fibers, respectively. However, the nanotube alignment is slightly lower than that of the PVA chains. It is about (9° for SWNT and (11° for MWNT fibers. The structural differences between untreated and hot-stretched fibers are coupled directly to distinct mechanical behavior. The strain versus stress curves shown in Figure 2 demonstrate that hot-stretched fibers exhibit a higher strength and, more importantly, a significantly greater energy absorption at lower strain. The fiber shown in Figure 2 exhibits a strain-to-failure of about 11% and a toughness of about 55 J/g, which is significantly higher than the toughness of Kevlar.8 We believe that the improvements of the strength and Young’s modulus are due primarily to the best alignment of the nanotubes and to the crystallinity of the PVA, which
Figure 4. Example of angular scattered intensity at Q ) 1.4 Å-1 (wave vector corresponding to the diffraction from the PVA chains as detailed in refs 12 and 13). (a) For a super-tough SWNT fiber. Assuming a Gaussian distribution for the chain orientation, as detailed in previous references (12 and 13), we deduce that the alignment is about (27° ((HWHM: half width at half-maximum of the Gaussian distribution). (b) For a SWNT hot-stretched fiber, the alignment is significantly more pronounced. A fit with a Gaussian distribution yields an orientation of (4.3°.
Figure 5. Comparison of the behavior of a super-tough (up) and a hot-stretched (bottom) fiber in water. (a) Initial fibers in the absence of water. Their diameter is about 30 µm. (b) 10 s after immersion in distilled water. Swelling of the super-tough fiber is observed. (c) 40 s after immersion. The hot-stretched fiber remains unchanged. 2214
Nano Lett., Vol. 5, No. 11, 2005
enhances the stress transfer between the polymer and the nanotubes.11 Finally, we also point out that crystalline PVA does not dissolve in water at room temperature. As a result, hot-stretched fibers do not swell and preserve their mechanical properties in water or humid conditions in contrast to super-tough-type fibers. This effect is evidenced in Figure 5, where the responses to water immersion of two fibers are compared. Untreated fibers swell readily, whereas hotstretched fibers do not exhibit any modifications, even after a long period of immersion in water. Even though the nanotube orientations reported in this work are to our knowledge the best achieved so far by a postsynthesis spinning process, the difference between the PVA and nanotube orientations suggests a limitation. This could be due to the intrinsic curvature and defects of nanotubes or to the fact that the fibers still contain entangled bundles of nanotubes,2 which may “lock” during the hotdrawing, preventing complete straightening and alignment of the microstructure. This means that improvements can still be expected by using straighter or less-bundled nanotubes. At least 1 order of magnitude is predicted for the Young’s modulus in the last degrees of alignment.14 We believe that the new treatments and properties described in this work broaden the rich spectrum of applications of nanotube fibers, which already includes actuators,9,15 microelectrodes,16 electronic textiles, and super-capacitors.8 In addition to their unique multifunctional character, the novel structures reported currently make nanotube fibers competitive with industrial high-performance fibers on a technical basis.10 Acknowledgment. This work has been done in the framework of the GDRE 2756 on Science and Applications of Nanotubes and is supported by the DGA.
Nano Lett., Vol. 5, No. 11, 2005
References (1) Gommans, H. H.; Alldredge, J. W.; Tashiro, H.; Park, J.; Magnuson, J.; Rinzler, A. G. J. Appl. Phys. 2000, 88, 2509. (2) Vigolo, B.; Pe´nicaud, A.; Coulon, C.; Sauder, C.; Pailler, R.; Journet, C.; Bernier, P.; Poulin, P. Science 2000, 290, 1331. (3) Zhu, H. W.; Xu, C. L.; Wu, D. H.; Wei, B. Q.; Vajtai, R.; Ajayan, P. M. Science 2002, 296, 884. (4) Jiang, K.; Li, Q.; Fan, S. Nature 2002, 419, 801. (5) Li, Y.; Kinloch, I. A.; Windle, A. H. Science 2004, 304, 276. (6) Ericson, L. M.; Fan, H.; Peng, H.; Davis, V. A.; Zhou, W.; Sulpizio, J.; Wang, Y.; Booker, R.; Vavro, J.; Guthy, C.; Parra-Vasquez, A. N. G.; Kim, M. J.; Ramesh, S.; Saini, R. K.; Kittrell, C.; Lavin, G.; Schmidt, H.; Adams, W. W.; Billups, W. E.; Pasquali, M.; Hwang, W.-F.; Hauge, R. H.; Fischer, J. E.; Smalley, R. E. Science 2004, 305, 1447. (7) Zhang, M.; Atkinson, K.; Baughman, R. H. Science 2005, 306, 13581361. (8) Dalton, A. B.; Collins, S.; Mun˜oz, E.; Razal, J. M.; Ebron, V. H.; Ferraris, J. P.; Coleman, J. N.; Kim, B. G.; Baughman, R. H. Nature 2003, 423, 703. (9) Baughman, R. H. Science 290, 1310 2000. (10) Fourne´ F. Synthetic Fibers: Machines and Equipment, Manufacture, Properties: Handbook for Plant Engineering, Machine Design, and Operation; Hanser Gardner Publisher, Munich, 1999. (11) Cadek, M.; Coleman. J. N.; Ryan, K. P.; Nicolosi, V.; Bister, g.; Fonseca, A.; Naggy, J. B.; Szostak, K.; Beguin, F.; Blau, W. T. Nano Lett. 2007, 4, 353-356. (12) Launois, P.; Maurcci, A.; Vigolo, B.; Bernier, P.; Derre, A.; Poulin, P. J. Nanosci. Nanotechnol. 2001, 1, 125-128. (13) Vigolo, B.; Poulin, P.; Lucas, M.; Launois, P.; Bernier, P. Appl. Phys. Lett. 2002, 81, 1210. (14) Liu, T.; Kumar, S. Nano Lett. 2003, 3, 647. (15) Vigolo B.; Launois, P.; Lucas, M.; Badaire, S.; Bernier, P.; Poulin, P. Mater. Res. Soc. Symp. Proc. 2002, 706. (16) Wang, J.; Deo, R. P.; Poulin, P.; Maugey, M. J. Am. Chem. Soc. 2003, 125, 14706.
NL051419W
2215
