Dispersing Carbon Nanotubes: Towards Molecular Understanding

Sammanfattning: Carbon nanotubes (CNTs) exhibit unique and fascinating intrinsic electrical, optical, thermal or mechanical properties that lead to a plethora of potential applications in composite materials, electronics, energy storage, medicine, among others. However, the manipulation of nanotubes is not trivial and there are significant difficulties to overcome before achieving their full potential in applications. Because of their high aspect ratio and strong tube-to-tube van der Waals interactions, nanotubes form bundles and ropes that are difficult to disperse in liquids. In this thesis, the topic of dispersing carbon nanotubes in water was addressed by several experimental methods such as nuclear magnetic resonance (NMR) diffusometry and light/electron microcopy. The main goal was to obtain molecular information on how the dispersants interact with carbon nanotubes.In dispersions of single-walled carbon nanotubes (SWNTs) in water, only a small fraction of the polymeric dispersant (Pluronic F127) was shown to be adsorbed at the CNT surface. Regarding dynamic features, the residence time of F127 on the SWNT surface was measured to be in the order of hundred milliseconds, and the lateral diffusion coefficient of the polymer along the nanotube surface proved to be an order of magnitude slower than that in the solution. The surface coverage of SWNTs by F127 was also investigated and the competitive adsorption of F127 and the protein bovine serum albumin, BSA, was assessed. F127 was found to bind stronger to the CNT surface than BSA does.Low molecular weight dispersants, viz. surfactants, were also investigated. Using carefully controlled conditions for the sonication and centrifugation steps, reproducible sigmoidal dispersibility curves were obtained, that exhibited an interesting variation with molecular properties of the surfactants. Various metrics that quantify the ability of different surfactants to disperse CNTs were obtained. In particular, the concentration of surfactant required to attain maximal dispersibility depends linearly on alkyl chain length, which indicates that the CNT-surfactant association, although hydrophobic in nature, is different from a micellization process. No correlation between dispersibility and the critical micellization concentration, cmc, of the surfactants was found. For gemini surfactants of the n-s-n type with spacer length s and hydrophobic tail length n, the dispersibility of multiwalled carbon nanotubes (MWNTs) also followed sigmoidal curves that were compared to those obtained with single-tailed homologues. The increase in spacer length caused an increase in the dispersion efficiency. The observations indicate a loose type of monolayer adsorption rather than the formation of micelle-like aggregates on the nanotube surface. With the future goal of embedding nanotubes in liquid crystal (LC) phases and thereby creating nanocomposites, the effect of the spacer length on the thermotropic behavior of the gemini 12-s-12 surfactant was investigated. Different mesophases were observed and a non-monotonic effect of the spacer length was found and rationalized within a model of the surfactant packing in the solid state.The relative binding strength of simple surfactants to CNTs was assessed by the amount of F127 they displace from the CNT surface upon addition. Anionic surfactants were found to replace more F127, which was interpreted as a sign of stronger binding to CNT. The data collected for all surfactants showed a good correlation with their critical dispersibility concentration that suggests the existence of a surface coverage threshold for dispersing nanotubes.On the macroscopic scale, the formation of weakly bound CNT aggregates in homogeneous dispersions was found to be induced by vortex-shaking. These aggregates could quickly and easily be re-dispersed by mild sonication. This counterintuitive behavior was related to the type of dispersant used and of the duration of mechanical agitation and was explained as a result of loose coverage by the dispersant.