Page 388 - Fiber Fracture

370 J. Bernholc et al.

Fig. 11. Time evolution of the (17,O) tube with an addimer under 7.5% strain at 3000 K, illustrating the
spontaneous winding of the defect about the tube: (a) initial configuration consisting of a single turn; (b)
final configuration corresponding to about 3 turns after 1 ns.

The winding of the defect about the nanotube suggests that the combination of
addimers plus strain in the 5-10% range may be a natural way to produce differ-
ent electronic heterojunctions, thereby leading to the formation of different carbon
nanotube-based quantum dots. To test this idea, we constructed various addimer-based
defect structures at different strains and annealed at temperatures in the 2500-3000 K
range, with the following results. Structures on the (10,lO) armchair tubes were not
observed to be stable. Competing bond rotations (e.g., bond rotations away from the
defect, or bonds on the heptagons) lead to the degradation of the structure within a
few nanoseconds. It therefore seems quite unlikely that good quantum dots can be
made with the help of addimers from the (10,lO) and/or other armchair tubes. Much
more promising results were obtained for the (17,O) zigzag tube. Fig. 11 shows sample
configurations of such a tube consisting of 682 atoms, annealed at 3000 K and 7.5%
strain. As is evident, there is no sign of any competing ductile behavior that would lead
to the degradation of the structure. Rather, hexagons are added in a uniform fashion
about the circumference of the tube, ultimately leading to the formation of two to three
different windings of an (8,8) tube over the period of a nanosecond. This suggests that
with the addition of addimers to strained zigzag tubes one can selectively induce ductile
behavior on tubes that are otherwise brittle, and thus form clean interfaces between
tubes of different helicities.
To characterize the electronic properties of the induced (17,0)/(8,8)/( 17,O) structure
we have calculated the local density of states (LDOS) using a recursion method
(Haydock et al., 1975) within a tight-binding description of the carbon n bonds. Only
nearest-neighbor interactions were considered. Depending on the length of the (8,8)
segment, a clear emergence of a quantum dot structure was observed, with orbitals
becoming localized in the (8,8) segment and falling inside the fundamental energy gap
of the (17,O) tube (Chico et al., 1998; Orlikowski et a]., 1999).