Small miracles--turning nanodreams into reality: high-performance composites from carbon nanotubes--an integrated approach for success.

[ILLUSTRATION OMITTED]

Composites are a class of material in which fillers or additives are combined with a matrix (e.g. polymer, ceramic, metal) to produce materials with properties significantly enhanced over those of the neat matrix. The enhancement could be for increased mechanical performance (e.g. strength, toughness, wear resistance), or improved electrical or thermal conductivities. Composite materials are used to make an enormous number of diverse products, ranging from car tires, in which carbon black or clay is added to improve wear resistance, to concrete, where sand and stone are added to improve strength, to carbon fibre laminates, where carbon fibres are added to polymer resins to produce high-strength lightweight structures.

The Holy Grail in composite science is to find a single additive to impart multiple functionalities on the final composite that is a single lightweight, high-strength, highly conductive, self-monitoring, and self-healing material. Carbon nanotubes (CNT), in principle, should be the ideal filler to impart all of these properties. Indeed, when taken on their own, CNT exhibit the highest mechanical, electrical, and thermal properties of any known material. Recent mechanical measurements report the Young's modulus and ultimate tensile strength as high as 3.3 TPa (1) and 63 GPa (2), respectively. In comparison, structural steel exhibits an ultimate strength of only 0.4 GPa and is 6 times heavier for a given volume. As electronic materials, CNT have displayed both extraordinary metallic characteristics, showing room temperature ballistic transport (3) and conducting more than 1,000 times more current per unit area than copper, (4) as well as extraordinary semiconducting characteristics. Additionally, CNT have shown among the highest thermal conductivities of any known material (as high as 3500 W/mK), (5) even greater than that of pure diamond. Combined with their very high aspect ratios (length/diameter) that can reach well over 10,000, CNT are truly the ultimate additives for the fabrication of multifunctional composites.

What are carbon nanotubes?

Carbon nanotubes resemble single sheets of graphite, known as graphene, that have been rolled upon themselves to form hollow, straw-like structures. They are divided into two general classes according to the thickness of the tube walls. Single-walled carbon nanotubes (SWNT) are composed of a single graphene wall whereas multi-walled carbon nanotubes (MWNT) are made of several concentric cylinders with a wall separation of 0.34 nanometres. It is generally agreed that SWNT are superior to MWNT for composite applications for a number of reasons, especially if multifunctionality is sought. Firstly, SWNT have a much lower percolation thresh-old, the minimum amount required to obtain a continuous network within a matrix. As a result, MWNT must be added at several times the loading of SWNT to achieve similar performance, which has important implications on processing and manufacturability. Second, SWNT have much higher aspect ratios than MWNT, an important aspect for efficient load transfer and toughening mechanisms. Third, MWNT typically possess greater numbers of defects than SWNT, which limits the maximum possible performance achievable by the composite. Fourth, SWNT are superior in combined properties to MWNT. For example, although SWNT and MWNT have very similar mechanical properties, SWNT have thermal conductivities that are at least one order of magnitude better than MWNT. (6) Finally, the properties of SWNT are strongly chirality dependent, creating the opportunity for tuneable performance once chirality-selective synthesis and/or separation are realized on a practical scale. This is not possible for MWNT as each individual graphene cylinder is independent from the others.

CNT composites today

The development of advanced composites based on CNT is still very much in its infancy. Despite more than 3,800 published scientific articles related to this field at the time of this writing, all CNT-based composites reported to date have shown poorer than expected performances, (2,7) and useful multifunctionality has yet to be demonstrated. The main reasons for this are: highly variable purity and quality of the CNT samples used; lack of standards for quality and purity assessment and; lack of effective chemistry for purification, dispersion/exfoliation and binding to the composite matrix. (7,8,9) Over much of the past decade, our team at the National Research Council Canada (NRC) has been working steadily to address each of these issues. This has led to the development of a very successful integrated approach to designing CNT composites in which we exercise control over each stage of the process from nanotube synthesis through processing and integration to composite fabrication and testing.

CNT supply

High variability in the purity and quality of CNT samples has been a problem plaguing the CNT-composites field since its very beginnings. In this context, "purity" refers to the fraction of the sample that is CNT, as opposed to impurity elements and other forms of carbon. "Quality" is a measure of the degree to which the CNT are free from defects. Part of the cause is that there are numerous methods by which CNT can be made, (10) each producing material of slightly different composition and properties. Purity can be improved by applying post-production purification techniques, but quality is largely dependent on the production process and is difficult to improve at a later time.

A consensus is rapidly forming in the community that quality is paramount to obtaining high-performance composite materials from CNT. The highest quality SWNT can be synthesized reliably at the laboratory scale using the laser vaporization technique. Our team at the NRC has developed a unique two-laser process (11) that has proven especially well suited to this task, and is being adopted by other groups around the world to produce exceptional quality SWNT. Unfortunately, this and other laser methods are not truly scalable, making them impractical for composites applications requiring large amounts of SWNT. Plasma discharge and chemical vapour decomposition (CVD) methods are more suited to large scale production but generally suffer from lower levels of material quality. Figure 1 illustrates the magnitude of the problem related to reliability in current SWNT supply. Raman spectra of the most prominent commercial materials are compared with NRC laser-grown SWNT, which we use as a standard. Without entering the complexity of Raman spectroscopy as it applies to SWNT, the features of note lie in the 150-250 [centimetre.sup.-1], 1250-1450 [centimetre.sup.-1], and 1500-1650 centi[metre.sup.-1] regions, which are attributed to the radial breathing mode (RBM), disorder-induced D-band, and graphitic G-band, respectively. The RBM is very specific to SWNT, its Raman shift being related to the SWNT diameter. If the RBM is not observed it is because there are no SWNT present, which is the case for one of the commercial products shown in the figure. The relative intensity of the D-band is indicative of the quality/purity of the sample, with lower intensity being better. It is clear from Figure 1 that the materials differ greatly in quality/purity, with the NRC laser-grown material standing out above the others. Despite these problems, enormous progress has been made in the past few years and it is only a matter of time until reliable low-cost sources of large quantities of high quality SWNT are available.

[FIGURE 1 OMITTED]

[FIGURE 2 OMITTED]

MWNT are already available at the kiloton level, but reliability in their quality is as much a problem as it is for SWNT. Recently, a Japanese company has introduced an in-situ annealing treatment during synthesis that has improved the quality of the MWNT significantly. Reliability in the purity of MWNT is much higher than SWNT, mainly because processes to produce MWNT have been in existence since the mid-1970s whereas SWNT were only discovered in 1993, and their growth conditions are much more stringent than for MWNT.

Standardization is essential

Compounding the problems associated with the variability in CNT supply is the fact that there are currently no internationally recognized standard practices for characterizing CNT material and reporting its composition and properties. Standards are an essential component for enabling the widespread development of commercial activities, especially by those with less expertise in the CNT field. Companies need to be certain that their raw materials will reliably perform as expected. This is an especially acute problem for CNT since it is difficult to distinguish them from the other forms of nano-structured carbon typically present as by-products of synthesis. Standards for CNT are actively being developed by the International Organization for Standardization (ISO) and Canada is actively playing a role through the NRC's Institute for National Measurement Standards. Until these standards are widely accepted, buyers of CNT should exercise due caution and should request as much certification data as possible.

Chemistry is the key

Up until very recently, the majority of efforts to make high-performance composites from CNT involved a simple physical mixing of the nanotubes into a matrix with the hope that some remarkable new composite would result. Unfortunately, little to no enhancement of composite properties was observed. The reason for this stems from the structure of the nanotubes themselves. CNT are fully aromatic with [sp.sup.2] hybridized sidewalls, giving them low chemical reactivity, very low solubility in most solvents, and weak affinity for most common composite matrices. In addition, just as graphene sheets prefer stacking to form graphite, CNT have a strong inter-tube attraction that causes them to bundle into thick rope-like structures (see Figure 2). The energy to de-bundle two SWNT, for example, is about 12 kilocalories (0.5 eV) per nanometre of length, which is a significant amount of energy considering SWNT can be several microns long. Therefore, to take full advantage of CNT in composite applications one must first overcome the bundling force in order to uniformly disperse the CNT as well as increase the interaction between the CNT and the matrix. Both of these goals can be reached through proper chemical modification of the CNT themselves, which is now recognized as the key to fully leveraging carbon nanotubes' remarkable properties. This chemistry can be accomplished in a number of ways, including wrapping the CNT with a polymer chain, non-covalent [pi]-[pi] stacking to the delocalized aromatic network, and direct covalent ([sp.sup.3]) functionalization to carbon atoms in the nanotubes. Our group at the NRC has developed extensive expertise in the covalent attachment of SWNT to various matrices through the linking of tailored functional moieties to SWNT sidewalls. Several strategies have been developed and proven that offer excellent flexibility and control over this process. One method we favour is to perform chemistry on reduced (negatively charged) SWNT because the nanotubes are naturally exfoliated during the process, and it substantially reduces the time and cost of the functionalization. (12) Figure 3 illustrates the effectiveness of our methods for CNT-composites. The first two panels show covalently functionalized SWNT dispersed in an epoxy resin, while the third shows the result of physical mixing unmodified SWNT.

[FIGURE 3 OMITTED]

An integrated approach for success

In order to counter the challenges discussed above and to discover what is ultimately possible with carbon nanotubes, we have adopted an integrated approach to the development of CNT-composites. In this approach we exert strict control over every step of the development process, including SWNT synthesis, purification, characterization, functionalization, and integration with the matrix. In this way we are able to exercise quality control and traceability at every stage. This approach is already proving successful with recent demonstration that the fracture toughness in epoxy resins can be improved by more than 60 percent with the addition of as little as 0.16 weight-percent of functionalized SWNT.

Future of CNT composites

There have been high expectations and many promises regarding CNT-composites, with few being fulfilled as yet. Fortunately, the CNT community is now realizing that reliability, quality, standardization, and chemistry are essential to making concerted progress. Through our experience we now know that with an integrated approach it is possible to make high-performance composite materials with CNT. Just how far the performance of these materials can be pushed remains to be seen. One prediction we can make is that the next few years will bring great advancements to the field and move us even closer to developing the ultimate multifunctional composites.

References

(1.) Kazuki Enomoto, Shintaro Kitakata, Toshiyuki Yasuhara, Naoto Ohtake, Toru Kuzumaki, and Yoshitaka Mitsuda, "Measurement of Young's Modulus of Carbon Nanotubes by Nanoprobe Manipulation in a transmission electron microscope," Applied Physics Letters 88 (April 2006) pp.153115-153117.

(2.) Min-Feng Yu, Oleg Lourie, Mark J. Dyer, Katerina Moloni, Thomas F. Kelley, and Rodney S. Ruoff, "Strength and breaking Mechanism of Multiwalled Carbon Naotubes Under Tensile Load," Science 87, (January 2000) pp. 637-640.

(3.) Philippe Poncharal, Claire Berger, Yah Yi, Z. L. Wang, and Walt A. de Heer, "Room Temperatura Ballistic Conduction in Carbon Nanotubes," Journal of Physical Chemistry B 106, 47 (November 2002) pp. 12104-12118.

(4.) B. Q. Wei, R. Vajtai, and P. M. Ajayan, "Reliability and Current Carrying Capacity of Carbon Nanotubes," Applied Physics Letters 79, 8 (2001) pp. 1172-1174.

(5.) Eric Pop, David Mann, Qian Wang, Kenneth Goodson, and Hongjie Dai, "Thermal Conductance of an Individual Single-wall Carbon Nanotube Above Room Temperature," Nanoletters 6, 1 (January 2006) pp. 96-100.

(6.) Tae-Youl Choi, Dimos Poulikakos, Joy Tharian, and Urs Sennhauser, "Measurement of the Thermal Conductivity of Individual Carbon Nanotubes by the Four-Point Three-[omega] Method," Nanoletters 6, 8 (August 2006) pp. 1589-1593.

(7.) Jonathan N. Coleman, Umar Khan, and Yurii K. Gun'ko, "Mechanical Reinforcement of Polymers Using Carbon Nanotubes," Advanced Materials 18, 6 (March 2006) pp. 689-706.

(8.) Jean-Paul Salvetat, Sanjib Bhattacharyya, and R. Byron Pipes, "Progress on Mechanics of Carbon Nanotubes and Derived Materials," Journal of Nanoscience and Nanotechnology 6, 7 (2006) pp. 1857-1882.

(9.) Erik T. Thostenson, Chunyu Li, and TsuWei Chou, "Nanocomposites in context," Composites Science and Technology 65, 3-4 (March 2005) pp. 495-516.

(10.) Christopher T. Kingston and Benoit Simard, Analytical Letters 36, 15 (2003) pp. 3139-3145.

(11.) Christopher. T. Kingston, Zygmunt J. Jakubek, Stephane Denommee and Benoit Simard, Carbon 42, 8-9 (2004) pp. 1657-1664.

(12.) Yadienka Martinez-Rubi, Jingwen Guan, Shuqiong Lin, Christine Scriver, Ralph E. Sturgeon and Benoit Simard, "Rapid and Controlabte Covalent Functionalization of Single-Walled Carbon Nanotubes at Room Temperature," Chemical Communication, 2007, DOI: 10.1039/b712299c.

Stephane Denommee, MCIC, is a technical officer with expertise with nanomaterials.

Jingwen Guan is a research officer with expertise in the chemistry of SWNT

Christopher Kingston is a research officer with expertise in the synthesis and characterization of SWNT.

Yadienka Martinez-Rubi is an NSERC post-doctoral fellow with expertise in the chemistry of SWNT

Benoit Simard, FCIC, is principal research officer and group leader of the Molecular and Nano-Material Architectures Group at the National Research Council Canada's Steacie Institute for Molecular Sciences (SIMS).