Tensile Tests on Individual Single‐Walled Carbon Nanotubes: Linking Nanotube Strength with Its Defects

www.MaterialsViews.com C O M M U Tensile Tests on Individual Single-Walled Carbon Nanotubes: Linking Nanotube Strength with Its Defects N IC A By Ming-Sheng Wang , * Dmitri Golberg , * and Yoshio Bando IO N Superb toughness of CNTs, seamless cylinders made of graphene sheets, has attracted prime attention of mechanical engineers. Theoretical calculations predict that an ideal single-walled carbon nanotube (SWNT) possesses an extremely high tensile strength of 75–135 GPa, depending on tube chirality. [ 1–3 ] However, most previous experimental results from several groups have typically been less sound. [ 4–7 ] Taking Yu et al.’s report as an example, tensile tests on 19 individual multi-walled nanotubes (MWNTs) pointed at the failure strengths of 11–63 GPa, with a mean value of only 28 GPa. [ 4 ] These theory-experiment discrepancies actually arise from the presence of structural defects in nanotubes introduced during their growth and/or post-growth treatments. Although the effects of such defects on tube fracture have only been computed via e.g. quantum mechanical or molecular mechanical calculations, [ 2 ] an experimental study clearly linking the nanotube defects to its fracture strength has not been available to date. Furthermore, the processes of defect nucleation, their propagation and relations to the tube failure at loading have also remained elusive and untested. We also noted that till now, only a few sets of CNT fracture measurements have been reported for any tube structure, either individual MWNTs or ropes of SWNTs. In addition, experimental data on such structures suffers from various uncertainties arising from e.g. interlayer (or intra-tube) load transfer, layer and/or tube sliding, and the estimation of an effective cross-sectional area. Also, the outermost layer of MWNTs usually has a much larger diameter than the model SWNTs used in most of theoretical calculations. This discrepancy brings other complications for the direct comparison between experimental results and theoretical simulations. The only direct and most reliable way is to take an individual SWNT and to conduct a direct strength measurement on it. Such experiment would be highly desirable for both experimentalists and theoreticians within the Nanotube community. [ 8 ] However, the extremely small dimensions, e.g. a diameter of a few nanometers, and a micrometer scale length impose a tremendous burden on an experimentalist. The diffi culties include visualizing, picking and placing of such tiny, fl exible individual SWNTs followed by nanoclamp fabrication within a force-sensor microdevice. Although some mechanical