Microscopic investigation of laser-induced structural changes in single-wall carbon nanotubes

Extensive excited-state molecular dynamics simulations of femtosecond laser-induced structural transformation in single-walled carbon nanotubes (SWNTs) are presented. We considered in the simulation two limiting cases; one where only a short portion of the tube is irradiated and the other where the whole tube is affected by the laser pulse. We have analyzed the role of chirality (zigzag versus armchair and some chiral tube cases) in the damage threshold as a function of tube diameter. Nontrivial dependence of the damage threshold as a function of diameter has been found. We find that for equal laser parameters, zigzag SWNTs are on average equally stable with respect to laser excitation as armchair SWNTs, but their stabilities show different dependencies on diameter. Due to the higher stiffness of the $(n,n)$ tubes in the direction perpendicular to its axis as compared to the $(n,0)$ tubes, we find the formation of standing waves in the nanotube wall for zigzag and not for armchair tubes. We also studied the role of the laser pulse duration and show that in general longer laser pulses increase the damage threshold. This result is rationalized in terms of electron-ion relaxation times. Implications of laser-induced structural transformations are analyzed.

Figure 1

(Color online) Schematic illustration of the two different types of boundary conditions used in the MD simulations on the nanotubes.

Figure 2

(Color online) Damage thresholds of zigzag (a) and armchair (b) SWNTs as a function of tube diameter. The damage threshold is defined as the minimum energy deposited by the laser in order to produce an irreversible structural transformation of the tube. Both panels correspond to fixed periodic boundary conditions. This corresponds to long nanotubes that are only partially excited by the laser pulse and thus cannot expand in the longitudinal direction. Note that for zigzag tubes, the duration of the laser pulse was $\tau =20\mathrm{fs}$, whereas it was $\tau =50\mathrm{fs}$ for armchair tubes.

Figure 3

(Color online) Snapshots of the time evolution of a (22,0) SWNT after laser excitation 30% above the damage threshold. The duration of the laser pulse was $\tau =20\mathrm{fs}$, and the deposited energy $E_0=2.45\mathrm{eV}$/atom. Longitudinal expansion of the tube upon laser excitation was not permitted by fixing the boundary conditions.

Figure 4

(Color online) Damage thresholds of zigzag (a) and armchair (b) SWNTs as a function of tube diameter. Both panels correspond to variable periodic boundary conditions. This corresponds to short nanotubes that fit entirely into the laser spot and can thus expand freely upon laser excitation. Note that for zigzag tubes, the duration of the laser pulse was $\tau =20\mathrm{fs}$ whereas it was $\tau =50\mathrm{fs}$ for armchair tubes.

Figure 5

Damage thresholds of $(n,n)$ SWNTs as a function of tube diameter for variable tube length (see text). The duration of the laser pulses was varied from $\tau =20\mathrm{fs}$ (star symbols, dotted error bars) over $\tau =50\mathrm{fs}$ (cross symbols, dashed error bars) to $\tau =100\mathrm{fs}$ (diamond symbols, solid error bars).

Figure 6

(Color online) Damage thresholds for chiral tubes as a function of chiral angle for a diameter $d\approx 8\mathrm{\AA }$. Note that the variation of thresholds is smaller than in Figs. 2, 4. Variable boundary conditions allowed the tube length to fluctuate. The laser pulse duration was $\tau =20\mathrm{fs}$.

Figure 7

(Color online) Snapshots of the time evolution of a (20,0) SWNT after laser excitation slightly below the damage threshold. The duration of the laser pulse was $\tau =20\mathrm{fs}$, the deposited energy $E_0=2.1\mathrm{eV}$/atom. Longitudinal expansion of the tube upon laser excitation was permitted.

Figure 8

(Color online) Snapshots of the time evolution of a (20,0) SWNT after laser excitation slightly above the damage threshold. The duration of the laser pulse was $\tau =20\mathrm{fs}$, the deposited energy $E_0=2.35\mathrm{eV}$/atom. Longitudinal expansion of the tube upon laser excitation was permitted. Several atoms are emitted and leave the displayed area.

Figure 9

(Color online) Selected structure of the (22,0) nanotube with variable MD supercell excited just below the ablation threshold. Three examples of the standing wave, at $t=0$, 850, and $1940\mathrm{fs}$, respectively.

Figure 10

(Color online) Analysis of a (22,0) nanotube with variable MD supercell excited just below the ablation threshold. The radii of the sections oscillate in phase directly after laser excitation but after $2\mathrm{ps}$ some dephasing is seen. Each color corresponds to a section of the CNT as described in the text.

Figure 11

Absolute value of the (22,0) CNT deformation amplitude from Fig. 10, averaged over all sections of a CNT structure (for explanation see text), as a function of time.

Figure 12

(Color online) Orientation of the bonds in armchair (top) and zigzag (bottom) SWNTs with respect to the tube axis $\mathit{T}$. The four type of bonds are stretched to a different extent in the intense breathing modes vibrations excited by a laser pulse.

Figure 13

(a) Average bond length for the two kinds of bonds in an armchair (10,10) carbon nanotube, either perpendicular to the tube axis or not (see Fig. 12 for the definition of the bond types). The energy deposited by the $\tau =100\mathrm{fs}$ laser pulse is $E_0=2.17\mathrm{eV}$/atom (well below the damage threshold, see Fig. 5). (b) Average bond length for the two kinds of bonds in a zigzag (18,0) CNT, either along the tube axis or not. The energy absorbed from the $\tau =50\mathrm{fs}$ laser pulse was $E_0=2.02\mathrm{eV}$/atom (just below the damage threshold).