A latex-based concept for making carbon nanotube/polymer nanocomposites

Several methods have been developed over the last few years to achieve the incorporation of carbon nanotubes (CNTs) into a polymer matrix in order to obtain electrically conductive nanocomposites. The key factors for producing such composites with low CNT loadings comprise the quality of the wetting between the filler and the polymer matrix, as well as the state of dispersion of the CNTs throughout the matrix. The final target is to manufacture easilyprocessable, low density conductive plastics, that in the future would be able to replace metals in applications for which these are still preferred. Nevertheless, as produced CNTs are either stuck together in thick bundles in the case of singlewall carbon nanotubes (SWCNTs), or are highly entangled in the case of multi-wall carbon nanotubes (MWCNTs). As a consequence, one of the main bottlenecks for the production of high performance CNT/polymer nanocomposites remains the resistance of CNTs against individualization. The work presented in this Ph.D. thesis focuses on the study of: - the various steps of a process based on the application of latex technology, which is utilized to prepare CNT/polymer nanocomposites. The key step of this process is the mixing of two colloidal solutions, being a dispersion of mainly individual CNTs covered by surfactant molecules, and a polymer latex, i.e. an aqueous dispersion of submicron polymer particles. The latter are also covered with surfactant molecules. - the nanocomposites produced in this way, in particular their electrical properties, which are governed by the formation of a percolating network of CNTs, and are therefore strongly dependent on the concentration of conductive fillers dispersed in the matrix. The first step of the process consists of debundling CNTs in an aqueous surfactant solution (typically sodium dodecyl sulfate, SDS) in order to obtain a stable dispersion of CNTs covered by surfactant molecules. The achievement of this first step is crucial: in order to get conductive films with filler loadings as low as possible, it is imperative to achieve a good dispersion of the CNTs in the polymer matrix. Consequently, it is very important to control and to monitor the CNT debundling during the first step of the process. However, CNT debundling does not guarantee a proper dispersion of the CNTs in the final nanocomposite. It was demonstrated that this sonication-driven step can be monitored by UV-Vis spectroscopy, see Chapter 3. This method is based on the fact that individual CNTs absorb light in the wavelength region between 204 200 and 1200 nm. The debundling of SWCNTs or MWCNTs results in an increase of the concentration of individual CNTs, and finally in an increase of the UV-Vis signal. A leveling off of the UV-Vis absorbance, recorded as a function of time and/or the total amount of ultrasonic energy supplied to the system, indicates that the maximum degree of exfoliation has been achieved and that, accordingly, further energy input can be stopped in order to prevent unnecessary damage to the CNTs. These results were confirmed with cryo-Transmission Electron Microscopy (cryo-TEM) and Scanning Electron Microscopy (SEM). In addition, four different experimental techniques, based on UV-Vis spectroscopy, surface tension measurements, thermogravimetry and a modified version of the Maron’s titration, have been developed in order to determine the lowest amount of surfactant necessary to reach the highest degree of exfoliation of the CNTs (Chapter 4). The results obtained enabled us to estimate a lower limit of the specific surface area of exfoliated SWCNTs, as well as an estimation of the specific surface covered by one surfactant molecule when adsorbed on the SWCNT surface. It is worth mentioning that these procedures are in principle applicable to a large range of surfactant-particle systems. In a second step, the stable aqueous SDS-CNT dispersion is mixed with a polymer latex. The mixture obtained is freeze dried and subsequently compression-molded. Before melt processing, the system typically consists of closely-packed latex particles (polystyrene latex particles for our model system) mixed with CNTs, which are confined in the interstitial space between the polymer particles. Since flow of the polymer occurs during the compression molding step, CNTs can move through the polymer melt. As a result, the processing conditions have a large influence on the conductivity, as well as on the percolation threshold of the nanocomposite, see Chapter 5. According to SEM analysis, in the nanocomposite films obtained, the CNTs are homogeneously dispersed in the polymer matrix, and form a network of preponderantly individualized CNTs. In particular, electrically conductive nanocomposites with a percolation threshold of about 0.3 wt% (resp. 0.9 wt%) of SWCNTs (resp. MWCNTs) dispersed in a high molecular weight polystyrene produced by free radical emulsion polymerization can be obtained in this way. This latex-based process is extremely versatile since it enables us to disperse SWCNTs and MWCNTs into most of the polymers produced by emulsion polymerization or polymers which can be artificially brought into a latex form. Thanks to this method, CNTs were successfully dispersed in another amorphous polymer than polystyrene, viz. poly(methyl methacrylate), a semi-crystalline polymer, i.e. polypropylene, or in a polymer blend, namely a poly(2,6-dimethyl- 1,4-phenylene ether)/polystyrene blend (see Chapter 7). Additionally, several procedures were explored in order to improve the properties of the composites, i.e. to lower the percolation threshold and increase the conductivity. We demonstrated that tuning the characteristics of the polymer matrix (in particular its molecular weight distribution, see Chapter 5) is a promising way to optimize these properties. It was also shown that the choice of the polymer matrix itself is a relevant parameter (see Chapter 7). Also the characteristics of the CNTs themselves (among others: their type (SWCNTs or MWCNTs), their intrinsic conductivity, their degree of purity, their diameter, and their aspect ratio) are of importance. Summarizing, we can state that the latex-based production of electrically conductive nanocomposites leads to well-defined materials with preponderantly individual CNTs homogeneously dispersed in a polymer matrix. Interestingly, the electrical properties of these materials are strongly influenced by the characteristics of the polymer matrix and the CNTs chosen, as well as by the process parameters.