We report and provide justification for the consistently observed four experimental facts of the mass spectrometric data of carbon cluster ( ≥ 1) emission from the low-energy Cs + -irradiated single-walled carbon nanotubes (SWCNTs). Firstly, the diatomic carbon C 2 is the most abundant sputtered species for Cs + in the energy range 0.2 ≤ ( + ) ≤ 2.0 . Secondly, monatomic carbon C 1 is emitted with the least sputtering yield. Thirdly, at low cesium energies i.e., E(Cs + ) < 400 eV, the emitted species are C 2 , C 3 and C 4 . Lastly, as the irradiating Cs + energy increases, the normalized yield of C 1 monotonically increases while C 2 , C 3 and C 4 show gradual decrease and saturation. The experimental data for the normalized density of clusters and atoms = � / ∑ � follows the pattern 2 > 3 > 4 > 1 . Sputtering of clusters is proved here to be due to thermal spikes. Binary collision cascade theory does not explain cluster sputtering. A statistical thermal model is developed to explain the experimentally observed data.
2The probability of a cluster C x to be emitted is shown to be proportional to that for the creation of an x-member vacancy with formation energy at temperature T s as = { ( / ) + 1} −1 . The energies of formation of single and double vacancies 1 and 2 from DFT calculations and the ratio of normalized experimental yields ( 1 / 2 ) have been used to estimate T s in theFor given values of the formation energies and using the experimental ratios 1 / 2 we can determine the locally enhanced spike temperature . Once is evaluated, the formation energies of tri-and quartovacancies are obtained by using the ratios of normalized densities 2 / 3 and 2 / 4 . We show that by invoking thermal spikes, cluster emission from, and the multiple vacancy generation in, the Cs + −irradiated SWCNTs can be explained. We also suggest modifications toMonte Carlo type calculations of sputtering.