Ricardo Manuel Ferreira Fernandes

Abstract Debundling and dispersing single-walled carbon nanotubes (SWNTs) is essential for applications, but the process is not well understood. In this work, aqueous SWNT dispersions were produced by sonicating pristine SWNT powder in the presence of an amphiphilic triblock copolymer (Pluronic F127) as dispersant. Upon centrifugation, one obtains a supernatant with suspended individual tubes and thin bundles and a precipitate with large bundles (and impurities). In the supernatant, that constitutes the final dispersion, we determined the dispersed SWNT concentration by thermogravi-metric analysis (TGA) and UV-vis spectroscopy, and the dispersant concentration by NMR The fraction of dispersant adsorbed at the SWNT surface was obtained by H-1 diffusion NMR Sigmoidal dispersion curves recording the concentration of dispersed SWNTs as a function of supernatant dispersant concentration were obtained at different SWNT loadings and sonication times. As SWNT bundles are debundled into smaller and smaller ones, the essential role of the dispersant is to sufficiently quickly cover the freshly exposed surfaces created by shear forces induced during sonication. Primarily kinetic reasons are behind the need for dispersant concentrations required to reach a substantial SWNT concentration. Centrifugation sets the size threshold below which SWNT particles are retained in the dispersion and consequently determines the SWNT concentration as a function of sonication time.

Abstract When using amphiphilic polymers to exfoliate and disperse carbon nanotubes in water, the balance between the hydrophobic and hydrophilic moieties is critical and nontrivial. Here, we investigate the mode of surface attachment of a triblock copolymer, Pluronics F127, composed of a central hydrophobic polypropylene oxide block flanked by hydrophilic polyethylene oxide blocks, onto single-walled carbon nanotubes (SWNTs). Crucially, we analyze the composition in dispersant of both the as-obtained dispersion (the supernatant) and the precipitate-containing undispersed materials. For this, we combine the carefully obtained data from H-1 NMR peak intensities and self-diffusion and thermogravimetric analysis. The molecular motions behind the observed NMR features are clarified. We find that the hydrophobic blocks attach to the dispersed SWNT surface and remain significantly immobilized leading to H-1 NMR signal loss. On the other hand, the hydrophilic blocks remain highly mobile and thus readily detectable by NMR. The dispersant is shown to possess significant block polydispersity that has a large effect on dispersibility. Polymers with large hydrophobic blocks adsorb on the surface of the carbonaceous particles that precipitate, indicating that although a larger hydrophobic block is good for enhancing adsorption, it may be less effective in dispersing the tubes. A model is also proposed that consistently explains our observations in SWNT dispersions and some contradicting findings obtained previously in carbon nanohorn dispersions. Overall, our findings help elucidating the molecular picture of the dispersion process for SWNTs and are of interest when looking for more effective (i.e., well-balanced) polymeric dispersants.