required for various electrical applications without compromising the host
polymer's other preferred physical ... For example, it works better for polyimide5.
APPLIED PHYSICS LETTERS
VOLUME 83, NUMBER 14
6 OCTOBER 2003
Homogeneous carbon nanotubeÕpolymer composites for electrical applications Rajagopal Ramasubramaniama) and Jian Chenb) Zyvex Corporation, 1321 North Plano Road, Richardson, Texas 75081
Haiying Liu Department of Chemistry, Michigan Technological University, Houghton, Michigan 49931
共Received 11 June 2003; accepted 13 August 2003兲 Homogeneous carbon nanotube/polymer composites were fabricated using noncovalently functionalized, soluble single-walled carbon nanotubes 共SWNTs兲. These composites showed dramatic improvements in the electrical conductivity with very low percolation threshold 共0.05–0.1 wt % of SWNT loading兲. By significantly improving the dispersion of SWNTs in commercial polymers, we show that only very low SWNT loading is needed to achieve the conductivity levels required for various electrical applications without compromising the host polymer’s other preferred physical properties and processability. In contrast to previous techniques, our method is applicable to various host polymers and does not require lengthy sonication. © 2003 American Institute of Physics. 关DOI: 10.1063/1.1616976兴
specific polymer. For example, it works better for polyimide5 than polystyrene.6 We have recently developed a versatile, nondamaging controlled chemistry to functionalize carbon nanotube surfaces while preserving nearly all of the nanotube’s intrinsic properties.3 This noncovalent chemistry can provide nanotube solubilization in organic solvents that allows homogeneous dispersion of nanotubes in the host polymer matrix. This approach avoids the lengthy sonication and is applicable to many organic soluble polymers such as polycarbonate, polystyrene, poly共methyl methacrylate兲, etc. SWNTs produced by the high pressure carbon monoxide 共HiPco兲 process were purchased from Carbon Nanotechnologies, Inc., and were solubilized in chloroform with poly共phenyleneethynylene兲s 共PPE兲 along with vigorous shaking and/or short bath sonication.3 The mass ratio of PPE:SWNTs was kept at 0.5. The resulting PPE-functionalized SWNT solution was then mixed with a host polymer 共polycarbonate or polystyrene兲 solution in chloroform to produce a homogeneous nanotube/polymer composite solution. The uniform composite film was prepared from this solution on a silicon wafer with a 100 nm thick thermal oxide layer either by drop casting or by slow-speed spin coating. The sample was then heated to 80– 90 °C to remove residual solvent. Nanotube polymer composite films with various SWNT loadings from 0.01 to 10 wt % in polystyrene as well as in polycarbonate were prepared according to the above procedure. The SWNT loading values for PPE-functionalized SWNTs/host polymer composites quoted are based on pristine SWNT material only, and exclude the PPE material. We investigated the electrical conductivities of two different types of polymer composite: 共a兲 PPE-functionalized SWNTs/polystyrene 共PPE-SWNTs/polystyrene兲 composites and 共b兲 PPE-SWNTs/polycarbonate composites. The thicknesses of the films were measured using a LEO 1530 scanning electron microscope or a profilameter. The typical thicknesses of the composite films were in the range of 2–10 m. Electrical conductivity measurements were performed using
Carbon nanotubes, due to their high-aspect ratio, small diameter, light weight, high-mechanical strength, highelectrical and thermal conductivity, high-thermal and air stability, are recognized as the ultimate carbon fibers for high performance, multifunctional composites.1,2 However, smooth carbon nanotube surfaces 共i.e., sidewalls兲 are incompatible with most solvents and polymers which results in poor dispersion of nanotubes in the polymer matrix. We report here that homogeneous nanotube polymer composites can be fabricated using noncovalently functionalized, soluble single-walled carbon nanotubes 共SWNTs兲,3 and that these composites show dramatic improvements in electrical conductivity with low percolation threshold 共0.05–0.1 wt % of SWNT loading兲. By significantly improving the dispersion of SWNTs in commercial polymers, we show that only a very small amount of SWNTs are needed to achieve the conductivity levels required for different electrical applications without compromising the host polymer’s other desired physical properties and processability. The electrically conductive carbon nanotube/polymer composites4 – 8 should find various applications such as in electrostatic dissipation, electromagnetic interference 共EMI兲 shielding, printable circuit wiring, and transparent conductive coatings. Pristine SWNTs are generally insoluble in common solvents and polymers, and difficult to chemically functionalize without altering the nanotube’s desirable intrinsic properties. Two common approaches have been used previously to disperse SWNTs in a host polymer: 共1兲 dispersing the SWNTs in a polymer solution by lengthy sonication 共up to 48 h兲.2 The lengthy sonication, however, can damage/cut the SWNTs, which is undesirable for many applications. 共2兲 In situ polymerization in the presence of SWNTs.5,6 The efficiency of this approach, however, is highly dependent on the a兲
Electronic mail:
[email protected] Electronic mail:
[email protected]
b兲
0003-6951/2003/83(14)/2928/3/$20.00
2928
© 2003 American Institute of Physics
Appl. Phys. Lett., Vol. 83, No. 14, 6 October 2003
Ramasubramaniam, Chen, and Liu
2929
FIG. 2. 共a兲 Room temperature electrical conductivity of PPE-SWNTs/ polystyrene composite vs the SWNT weight loading. Dashed lines represent the approximate conductivity lower bound required for several electrical applications. 共b兲 Room temperature conductivity of the PPE-SWNTs/ polystyrene composite as a function of the reduced mass fraction of SWNTs.
about 10⫺14 S/m. 9 The conductivity of pristine 共unfunctionalized兲 HiPco-SWNT buckypaper is about 5.1⫻104 S/m. The conductivity of the composite increases sharply between 0.02 and 0.05 wt % SWNT loading, indicating the formation of a percolating network. At the onset of the percolating network, the electrical conductivity obeys the power law relation10
c⬀共 v ⫺ v c 兲,
FIG. 1. 共a兲 Surface and 共b兲 cross-sectional SEM images of PPE-SWNTs 共5 wt % SWNTs兲/polystyrene composite film. 共c兲 Closeup of 共b兲.
the standard four-point probe method to reduce the effects of contact resistance. A Philips DM 2812 power supply and a Keithly 2002 digital multimeter were used to measure the IV characteristics of the samples. Scanning electron microscope 共SEM兲 images of the surface and cross section of the PPE -SWNTs 共5 wt % of SWNTs兲/polystyrene composite film show the excellent dispersion of PPE-functionalized SWNTs in host polymer matrix. SWNTs are randomly distributed not only along the surface 关Fig. 1共a兲兴, but also throughout the cross section 关Figs. 1共b兲 and 1共c兲兴, indicating the formation of an isotropic, three-dimensional nanotube network in the host polymer matrix. This is essential for obtaining composites with isotropic electrical conductivity. Figure 2共a兲 shows the measured volume conductivity of PPE-SWNTs/polystyrene composites as a function of the SWNT loading. The conductivity of pure polystyrene is
共1兲
where c is the composite conductivity, v is the SWNT volume fraction, v c is the percolation threshold and  is the critical exponent. Since the densities of the polymer and the SWNT are similar, we assume that the mass fraction m and volume fraction v of the SWNT in the polymer are the same. As shown in Fig. 2共b兲, the PPE-SWNTs/polystyrene conductivity agrees very well with the percolation behavior given in Eq. 共1兲. The straight line with m c ⫽0.045% and  ⫽1.54 gives an excellent fit to the data with a correlation factor of 0.994, indicating an extremely low percolation threshold at 0.045 wt % SWNT loading. The very low percolation threshold is the signature of the excellent dispersion of high aspect ratio soluble SWNTs. Apart from the very low percolation threshold, we also observe that the conductivity reaches 6.89 S/m at 7 wt % SWNT loading, which is 14 orders of magnitude higher than that (10⫺14 S/m) of pure polystyrene, and 5 orders of magnitude higher than that (1.34⫻10⫺5 S/m) of the SWNTs 共8.5 wt %兲/polystyrene composite that was prepared by in situ polymerization.6 Figure 3共a兲 shows the measured volume conductivity of PPE-SWNTs/polycarbonate composites as a function of the SWNT loading. The conductivity of pure polycarbonate is
2930
Appl. Phys. Lett., Vol. 83, No. 14, 6 October 2003
FIG. 3. 共a兲 Room temperature electrical conductivity of PPE-SWNTs/ polycarbonate composite vs the SWNT weight loading. Dashed lines represent the approximate conductivity lower bound required for several electrical applications. 共b兲 Room temperature conductivity of PPE-SWNTs/ polycarbonate composite as a function of the reduced mass fraction of SWNTs.
Ramasubramaniam, Chen, and Liu
SWNT loading is required to achieve the necessary conductivity levels, the host polymer’s other preferred physical properties and processability would be minimally compromised. Blanchet et al. have shown that the percolative behavior of composites containing nanotubes in a conducting 共polyaniline兲 and in a nonconducting 共ethylcellulose兲 matrix is dramatically different.8 The good dispersion of SWNTs in doped polyaniline enables a low percolation threshold at 0.25 wt % SWNT loading, whereas the poor dispersion of SWNTs in ethylcellulose leads to a high percolation threshold at 3.0 wt % SWNT loading. The advantage of our approach is that it allows the homogeneous dispersion of SWNTs in various polymer matrices. In conclusion, we have demonstrated that homogeneous nanotube polymer composites can be fabricated using noncovalently functionalized, soluble SWNTs. These composites show dramatic improvements in the electrical conductivity with very low percolation threshold 共0.05–0.1 wt % SWNT loading兲, and should find various electrical applications. In addition, these composites can be used for mechanical, thermal, sensing and actuating applications due to the multifunctional nature of SWNTs. In contrast to previous techniques,2,5,6 our method is applicable to various host polymers and does not require lengthy sonication. The authors would like to thank Rishi Gupta, Keith Bradshaw, and John Goodnight for experimental assistance. 1
about 10⫺13 S/m. 9 We find that the conductivity of PPESWNTs/polycarbonate is generally higher that that of PPESWNTs/polystyrene at the same SWNT loading. For example, the conductivity reaches 4.81⫻102 S/m at 7 wt % SWNT loading, which is 15 orders of magnitude higher than that (10⫺13 S/m) of pure polycarbonate. As shown in Fig. 3共b兲, we also observe very low percolation threshold at 0.11 wt % SWNT loading (m c ⫽0.11%;  ⫽2.79). Figures 2共a兲 and 3共a兲 also show conductivity levels necessary for electrical applications such as electrostatic dissipation, electrostatic painting and EMI shielding.11 As shown in Fig. 3共a兲, 0.3 wt % SWNT loading in polycarbonate is sufficient for applications such as electrostatic dissipation and electrostatic painting, and 3 wt % SWNT loading is adequate for EMI shielding applications. Since only a very low
R. Andrews, D. Jacques, M. Minot, and T. Rantell, Macromol. Mater. Eng. 287, 395 共2002兲. 2 M. J. Biercuk, M. C. Llaguno, M. Radosavljevic, J. K. Hyun, A. T. Johnson, and J. E. Fischer, Appl. Phys. Lett. 80, 2767 共2002兲. 3 J. Chen, H. Liu, W. A. Weimer, M. D. Halls, D. H. Waldeck, and G. C. Walker, J. Am. Chem. Soc. 124, 9034 共2002兲. 4 B. E. Kilbride, J. N. Coleman, J. Fraysse, P. Fournet, M. Cadek, A. Drury, S. Hutzler, S. Roth, and W. J. Blau, J. Appl. Phys. 92, 4024 共2002兲. 5 C. Park, Z. Ounaies, K. A. Watson, R. E. Crooks, J. Smith, Jr., S. E. Lowther, J. W. Connell, E. J. Siochi, J. S. Harrison, and T. L. St. Clair, Chem. Phys. Lett. 364, 303 共2002兲. 6 H. J. Barraza, F. Pompeo, E. A. O’Rear, and D. E. Resasco, Nano Lett. 2, 797 共2002兲. 7 P. Po¨tschke, T. D. Fornes, and D. R. Paul, Polymer 43, 3247 共2002兲. 8 G. B. Blanchet, C. R. Fincher, and F. Gao, Appl. Phys. Lett. 82, 1290 共2003兲. 9 C. A. Harper, Handbook of Plastics, Elastomers, and Composites, 4th ed. 共McGraw–Hill, New York, 2002兲. 10 E. J. Garboczi, K. A. Snyder, J. E. Dongks, and M. F. Thorpe, Phys. Rev. E 52, 819 共1995兲. 11 B. Miller, Plastics World 54, 73 共1996兲.
