A series of pyrene-labeled dendrimers were prepared from generation n = 1 to n = 4 where the pyrenes were attached to the end groups of the dendrimers. Pyrene excimer formation was monitored by steady-state and time-resolved fluorescence spectroscopy as a function of generation number and in terms of the I(E)/I(M) ratio and the average rate constant of excimer formation . To account for the unconventional distribution of pyrene labels which were neither randomly distributed throughout the macromolecule nor limited to just two units which are the only two pyrene-labeling schemes that can be dealt with in a straightforward manner, a Model Free (MF) analysis was applied to the global analysis of the fluorescence decays. Within experimental error, the I(E)/I(M) ratios and obtained from, respectively, the steady-state fluorescence spectra and the time-resolved fluorescence decays were found to increase linearly with increasing generation number. This result is inconsistent with the fact that both the I(E)/I(M) ratio and are proportional to the local concentration of pyrene inside the dendrimer ([Py](loc)) which is not expected to increase with generation number if the excited pyrene is assumed to diffuse freely throughout the dendrimer interior. Since the core-dense model predicts that the dendrimer terminal ends can occupy any position throughout the dendrimer interior, these results suggest that excimer formation between the pyrene-labeled ends is enhanced due to the branched nature of the dendrimer.

The effect of pyrene distribution within pyrene‐functionalized random and block copolymers on noncovalent polymer/single‐walled carbon nanotube (SWNT) interactions was investigated. The block copolymers served as superior solubilizing agents in comparison with the random copolymers. Also, increasing the pyrene content within a polymer, while a constant molecular weight was maintained, improved SWNT solubility and therefore had to result in stronger polymer–nanotube interactions. However, increasing the length of the pyrene‐containing block diminished nanotube solubility, likely because of a lower number of polymer chains that were capable of binding to the nanotube surface. Atomic force microscopy and transmission electron microscopy indicated that the polymer–SWNT interactions were capable of partially debundling the nanotubes into individual solvated structures. © 2006 Wiley Periodicals, Inc. J Polym Sci Part A: Polym Chem 44: 1941–1951, 2006