A fundamental understanding of the mechanisms involved in the surfactant-assisted exfoliation and dispersion of carbon nanotubes (CNTs) in water calls for well-controlled experimental methodologies and reliable comparative metrics. We have assessed the ability of several ionic surfactants to disperse single and multiwalled carbon nanotubes, resorting to a stringently controlled sonication-centrifugation method for the preparation of the dispersions. The CNT concentration was accurately measured for a wide range of surfactant concentration, using combined thermogravimetric analysis and UV-vis spectroscopy. The obtained dispersibility curves yield several quantitative parameters, which in turn allow for the effects of nanotube morphology and surfactant properties (aromatic rings, chain length, headgroup charge, and cmc) to be assessed and rationalized, both in terms of dispersed nanotube mass and surface area. The data also indicate that the CNT-surfactant association follows patterns that are markedly different from other equilibrium processes governed by hydrophobicity (such as micellization); in particular, the surfactant concentration needed for maximum dispersibility, c(s,max), and the number of surfactant molecules per unit CNT area at c(s,max) are shown to depend linearly on chain length. The results further suggest that the presence of micelles in the exfoliation process is not a key factor either for starting CNT dispersibility or attaining its saturation value.

Noncovalent dispersion of carbon nanotubes is essential to most applications but still poorly understood at the molecular level. The interaction of the dispersing molecule with the nanotube, wrapping or nonwrapping, still awaits consensus. Herein, we have studied by 1 H NMR diffusometry some features of molecular dynamics in the system of carbon nanotubes dispersed by triblock copolymer Pluronics F127 in water. The diffusional decays obtained at different diffusion times, Δ, are not single-exponential and have a complex Δdependent profile, ultimately implying that the polymer is observed in two states: free (in unimeric form) and nanotubebound. Fitting a two-site exchange model to the data indicates that at any instant, only a small fraction of polymers are adsorbed on the nanotubes, with polydisperse residence times in the range of 100−400 ms. Most significantly, we further provide an estimate of D = (3−8) × 10 −12 m 2 s −1 for the coefficient of lateral diffusion of the polymer along the nanotube surface, which is an order of magnitude slower than the corresponding self-diffusion coefficient in water. The emerging picture is that of a nonwrapping mode for the polymer−nanotube interaction.

The binding strength of dispersants to the surface of carbon nanotubes is of crucial importance for the efficiency of the dispersion process and for potential applications, yet data are scarce on this subject. Here we present the results of diffusion NMR experiments in dispersions of single-walled carbon nanotubes (SWNTs) prepared by either the polymer Pluronics F127 or the protein bovine serum albumin (BSA). The experiments detect the amount of F127 molecules adsorbed onto the SWNT surface. This quantity is recorded (i) in F127-SWNT dispersions to which BSA molecules are added and (ii) in BSA-SWNT dispersions to which F127 molecules are added. The data clearly show that F127 replaces BSA adsorbed at the SWNT surface, while BSA leaves the adsorbed F127 coverage intact. Consequently, F127 binds to the nanotube surface more strongly than BSA. Hence, we provide a way to categorize dispersants by adsorption strength. We also provide evidence showing that the nanotubes dispersed by BSA form loose aggregates where a large part of the surface is not in direct contact with the surrounding liquid. The results are discussed in relation to previous findings in the literature.