Cross-linking between shells can be imparted onto a tube via electron- or ion-irradiation. Bombarding the tube with Gallium ions or electrons in a TEM will knock carbon atoms out of the tube. If the vacancies in adjacent shells are near each other, the tubes will link together. This is a very simple method for altering the structure of the tube, and may be responsible for the clear improvement in performance of irradiated tubes.

Carbon nanotubes have shown extremely high performance characteristics under tensile loads. These characteristics have been observed both in our own experiments using the in-situ testing device and in various computational models.

In the picture above, five sets of load-displacement data are plotted for tubes which have received varying irradiation doses. Note the wide range of stiffnesses shown, from the lowest (Test 5) to the highest (Test 2). Note also the difference in the failure mechanisms as observed in-situ the TEM. In the case of Test 4, the irradiation dose was low, implying a low density of defects. Since there were few crosslinks, a limited number of shells broke independently of the inner shells, resulting in a "telescopic" failure. In the case of Test 2, the defect density was high, so all of the shells shared the load until they all broke at once, as it would in the case of a metal tube.

Crosslinking Defects

The energetics of crosslinking defects can be studied by quantum mechanics. As shown below, these are defects in the shells of a multiwalled nanotube which produce an energetically favorable configuration of atoms that results in a bond between shells of the tube. That is, if two atoms are removed from the tube, one from each shell, the remaining atoms will preferentially build a carbon "bridge" between the tubes.

When the two shells are bound together, tensile load can be transferred between the shells of the tube, which means that for the same stress, the crosslinked tube will have lower strain than a similar unlinked tube.

Simulation Methods

Carbon nanotubes are of interest in many applications because of their extreme aspect ratio. A tube that is tens of microns long can still be less than 20 nm in diameter. This can be a problem in atomistic simulations because even single-walled tubes can contain thousands or even millions of atoms. Simulations containing this many atoms cannot be completed in a reasonable amount of time on available computers. The choice of simulation method is therefore a very important decision.

At the electronic level, Density Functional Tight-Binding (DFTB) methods have been shown to provide good results in a reasonable runtime. DFTB is a quantum mechanical theory based on the electron density at any point in the molecule. Certain approximations and pre-computed integrals allow for fast results without much sacrifice in accuracy. We are using DFTB and also molecular mechanics based on the Modified Tersoff-Brenner (2nd generation) potential (MTB2). This latest method has shown to exhibit good agreement with DFTB so we are using molecular mechanics to complement quantum mechanics simulations and to examine atomic structures similar to those tested in-situ the TEM.

Personnel

• H.D. Espinosa (PI)
• M. Locascio (Graduate Student)

Collaborators

• G. Schatz (Chemistry, NU)
• P. Zapol (Argonne National Lab)

Selected Publications

• H.D. Espinosa, Y. Zhu and N. Moldovan. "Design and operation of a MEMS-based material testing system for nanomechanical characterization," To appear in Journal of Microelectromechanical Systems, 2006.

• H. D. Espinosa, Y. Zhu, B. Peng and O. Loh, "Nano-scale testing of nanowires and carbon nanotubes using a microelectromechanical system." Chapter from Advances in multiphysics simulation of MEMS and NEMS. Edited by N. Aluru, C. Cercignani, A. Frangi, and S. Mukherjee, Imperial College Press, to appear 2007.