We have studied the exciton recombination dynamics of individual (6,4) and (6,5) single-walled carbon nanotubes embedded in aqueous gels or deposited on glass surfaces. CoMoCat nanotubes systematically display short monoexponential photoluminescence (PL) decays presumably due to defects introduced during their synthesis. In contrast HiPco nanotubes can either display mono-or biexponential PL decays depending on the environmental conditions. Transition from bi-to monoexponential decays can be reproduced by a simple three level model taking into account defect-dependent nonradiative decay mechanisms.The photoluminescence (PL) properties of semiconducting single walled carbon nanotubes (SWNTs) have attracted much attention over the last years.1 These properties strongly depend on the structure of each nanotube 2 but also on extrinsic factors resulting for instance from synthesis or environmental factors.
3-7As a result, PL studies performed on ensemble of SWNTs are affected by inhomogeneities that hinder the development of a detailed understanding of the underlying mechanisms. Experiments on individual SWNTs remove part of this heterogeneity by allowing description of the characteristics of each SWNT in its particular environment. 8 Comparisons between values of a physical parameter extracted from different experimental reports are however difficult because, generally, the studied samples differ by synthesis methods or preparation procedures. For example PL decays performed on individual (6,4) and (6,5) nanotubes by different groups have resulted in distinct behaviors, ranging from very short monoexponential decays to longer biexponential ones. 5,7,[9][10][11] This paper aims at understanding this large disparity in PL decay behaviors. We experimentally confirm that PL decays of (6,5) and (6,4) nanotubes can exhibit either mono-or biexponential behaviors and show that these depend on synthesis methods and nanotube environment. These observations are explained by a previous simple three-level model 10 taking now into account the defect-dependent dominant nonradiative decay mechanisms proposed by Pereibenos et al.
12The SWNTs used in this study were either synthesized using HiPco or CoMoCat methods. The nanotubes were dispersed in aqueous suspensions of the anionic surfactants sodium deoxycholate (DOC). For observations, single-molecule wide-field and confocal PL microscopes were used to image SWNTs excited with a continuous wave laser. The SWNTs were immobilized in aqueous agarose gels (5 wt %) or spin-coated on surfaces. SWNTs concentration was kept well below 1 µm -3 such that bright individual nanotubes could be optically resolved. The PL of these bright tubes was sent to a spectrometer for further spectral identifications of individual (6,4) or (6,5) nanotubes. Typically, isolated SWNTs feature narrow PL lines from their bright excitonic state E 11 , with full width at half-maximum (fwhm) in the range of ∼17-22 nm. Following identification of these bright individual nanotubes, the excitation modality was switche...