The mechanism of defect creation in supported nanotubes, i.e., nanotubes on different substrates, has not yet been addressed. In this project, we study the ion irradiation of single-walled nanotubes on different substrates. For obtaining qualitative understanding of the physical processes involved, we considered two limiting cases: a heavy-atom metallic substrate and a light-atom substrate with covalent bonds between atoms. The former was chosen to be a platinum (111) surface, the latter a Bernal graphite (0001) surface. Noble metals and graphite have repeatedly been used as substrates in experimental studies on carbon nanotubes.

Fig.1. A typical illustration of the defect production mechanism upon Ar ion impact is presented in the figure below where a number of snapshots corresponding to different time moments are given.

The first snapshot (a) shows the original geometry before the impact. The motion direction of the impinging 500-eV Ar ion (circled) is designated by the arrow. The Ar ion usually creates a single vacancy (or multi-atom vacancy) in the uppermost part of the nanotube wall as well as primary carbon recoils (b). Then the recoils and Ar ion produce defects in the lower part of the nanotube (c), resulting in sputtering C atoms from the nanotube and producing a collision cascade in the substrate. A part of the recoils (or the Ar ion) is reflected back from the surface producing some extra damage in the nanotube (d). The characteristic time of these processes is about 0.1 ps. The ion impact and the sputtered substrate atoms also result in the development of a pressure wave in the nanotube (e) propagating to its ends from the impact point. The oscillations in the nanotube diminishes during much longer time scale (the actual time is governed by the cooling rate). The final configuration after the system has been cooled down to the zero temperature is presented in (f).

Visual analysis of the final atom positions indicated that, analogously to the case of suspended nanotube irradiation, single vacancies are the most prolific defects in nanotubes which appear after ion impact.Two-coordinated single adatoms on both external and internal sides of the nanotube walls were also common. Besides this, other complex defects were observed. Those were Stone-Wales defects associated with a 90-degree-rotation of a bond in the nanotube atom network, small amorphous regions, local distortions in the network due to incorporated Pt atoms, non-hexagonal rings, and even at times very peculiar defect configurations like C-atom chains formed inside the nanotube when displaced C atoms were left in the nanotube interior.

Fig. 2. Here we plot here the number of coordination defects as a function of ion energy. Open circles stand for the numbers of atoms with a coordination other than three, triangles for the numbers of sputtered C atoms. Open squares correspond to the numbers of two-coordinated atoms.

It can be seen from Fig. 2 that, if the energy of the incident ion is higher than the defect creation threshold energy (about 40 eV), the number of defects increases with the energy up to roughly 600 eV, then it remains practically constant. The reason for such behavior is that at low energies the damage production grows with energy, since there is more energy available for it. At higher ion energies, although defect production in the nanotube drops as the Ar-C nuclear collision cross section decreases, there are more reflected C recoils and Pt atoms sputtered from the substrate, all of which damage the nanotube. At very high energies (several keV), the ion penetrates into the substrate rather deep, which results in the drop in the sputtering yield and, respectively, in the nanotube damage due to this mechanism. The decrease in damage due to the diminution in the cross section and the damage enhancement due to the sputtered atoms approximately counterbalance each other at energies higher than 1 keV.

The annealing of defects gives rise to a substantial drop in the defect numbers. In Fig. 2 we also present the overall damage as a function of ion energy (full circles) as well as the number of two-coordinated atoms (full squares) after annealing. The number of sputtered C atoms (triangles) remained, of course, the same. Although about 40% of defects annealed, a substantial amount of defects remained in the system. The ratio of annealed defects can be larger at longer time-scales, but total annealing of defects is not possible since a number of atoms were sputtered from the nanotube.

You can find more about the production of defects in suspended nanotube bundles in our recent publications.