Dispersion of single-walled carbon nanotubes of narrow diameter distribution.

The dispersibility and bundle defoliation of single-walled carbon nanotubes (SWNTs) of small diameter (<1 nm) have been evaluated on CoMoCAT samples with narrow distribution of diameters. As previously observed by photoluminescence and Raman spectroscopy, the CoMoCAT sample exhibits a uniquely narrow distribution of (n,m) structures that remains unchanged after different dispersion conditions. This narrow distribution allowed us to develop a method for quantifying the dispersability of the samples from their optical absorption spectra in terms of two ratios: the "resonance ratio" and the "normalized width." The former is defined as the quotient of the resonant band area and its nonresonant background. The latter is defined as the ratio of the width of the band at half-height to the peak height on a spectrum that has been normalized at 900 nm, making this an intensive property, rather than varying with the path length. In this study of the CoMoCAT sample, we have used the S22 transition corresponding to the (6,5) nanotube to do these calculations, which is the most abundant species. These two ratios provide a quantitative tool to compare different dispersion parameters (time of sonication, degree of centrifugation, etc.) on the same type of sample. From this comparison, an optimal procedure that maximizes the spectral features was selected; this procedure allowed us to contrast various surfactants at different pH values and concentrations. Several surfactants were as good or even better than the one we have used in previous studies, dodecylbenesulfonic acid sodium salt (NaDDBS). Despite differences in their dispersion abilities, none of the surfactants investigated generated new features in the absorption spectra nor changed the distribution of nanotube types, which confirms that the high selectivity of the CoMoCAT sample is in the original sample rather than caused by selective suspension of specific (n,m) nanotubes.