Bundle Effects on the Optical Properties of SWNTs Fano Effect

Although the origin of the 1540 cm-1 Breit-Wigner-Fano peak is not well explained, the Fano peaks are relevant to the bundle effect which is discussed in this subsection. This idea can be explained by the Raman spectra observed for the Br2 doped SWNT sample. The frequency of the RBMs are shifted upon doping, and from this frequency shift the charge transfer of the electrons from the SWNTs to the Br2 molecules can be measured [61]. This charge transfer enhances the electrical conductivity whose temperature dependence shows metallic behavior [62]. When SWNTs made by the arc method with the NiY catalyst are used, the undoped SWNT sample exhibits the RBM features around 170cm"1. When the SWNTs are doped by Br2 molecules, new RBM peaks appear at around 240 cm"1 when the laser excitation energy is greater than 1.8 eV, as shown in Fig. 18a. When the Raman spectra for the fully Br2 doped sample are measured, new features at 260 cm"1 are observed, but the peak at 260 cm"1 disappears and a new peak at 240 cm"1 can be observed for laser excitation energy greater than 1.96 eV (see Fig. 18) when the sample chamber is evacuated at room temperature, and the spectra for the undoped SWNTs are observed showing RBM peaks around 170 cm"1. Since heating in vacuum up to 250° C is needed to remove the bromine completely, the evacuated sample at room temperature consists of a partially doped bundle and an easily undoped portion, which is identified with isolated SWNTs, not in bundles. Since the Fermi energy shifts downward in the acceptor-doped portion of the sample, no resonance Raman effect is expected in the excitation

Fig. 18. (a) Resonance Raman spectra for bromine doped SWNTs prepared using a NiY catalyst. The sample is evacuated after full doping at room temperature. An additional peak around 240 cm"1 can be seen for laser excitation energies greater than 1.96 eV. (b) (left scale) The optical density of the absorption spectra for pristine (undoped) SWNT samples and (right scale) the intensity ratio of the RBMs at ~240cm"1 appearing only in the doped samples to the RBM at ~180cm"1 for the undoped sample. The additional RBM peaks appear when the metallic window is satisfied [63]

Fig. 18. (a) Resonance Raman spectra for bromine doped SWNTs prepared using a NiY catalyst. The sample is evacuated after full doping at room temperature. An additional peak around 240 cm"1 can be seen for laser excitation energies greater than 1.96 eV. (b) (left scale) The optical density of the absorption spectra for pristine (undoped) SWNT samples and (right scale) the intensity ratio of the RBMs at ~240cm"1 appearing only in the doped samples to the RBM at ~180cm"1 for the undoped sample. The additional RBM peaks appear when the metallic window is satisfied [63]

energy range corresponding to the semiconductor first and second peaks and the metallic third peak in the optical absorption spectra. In fact, in Fig. 18b, the intensity ratio of the Raman peaks around 240 cm-1 to that at 180 cm-1 is plotted by solid circles and the curve connecting these points is shown in the figure as a function of laser excitation energy. Also shown in the figure is the corresponding optical absorption spectrum for the pristine (undoped) sample plotted by the dotted curve. The onset energy of the Raman peaks at 240 cm-1 is consistent with the energy 2A EF which corresponds to the energy of the third metallic peak of the optical absorption. In fact, the optical absorption of the three peaks disappear upon Br2 doping (Fig.12) [24,41]. The peaks of Raman intensity at 240 cm-1 are relevant to resonant Raman scattering associated with the fourth or the fifth broad peaks of doped semiconductor SWNTs.

Raman shift (cm 1)

Fig. 19. The Raman Spectra for the undoped sample (top) and for the evacuated sample (bottom) after full Br2 doping at room temperature. The Fano spectral feature at 1540 cm-1 is missing in the spectrum for the evacuated sample [63]

Raman shift (cm 1)

Fig. 19. The Raman Spectra for the undoped sample (top) and for the evacuated sample (bottom) after full Br2 doping at room temperature. The Fano spectral feature at 1540 cm-1 is missing in the spectrum for the evacuated sample [63]

For this evacuated sample, the G-band spectra with the laser energy 1.78 eV is shown in Fig. 19. This laser energy corresponds to an energy in the metallic window, but no resonance Raman effect is expected from the doped bundle portion, as discussed above. Thus the resonant Raman spectra should be observed only in metallic nanotubes in the undoped portion of the sample which is considered to contain only isolated SWNTs. Surprisingly there are no 1540 cm-1 Fano-peaks for such an evacuated sample, although the un-doped sample has a mixture 1590cm-1 and 1540cm-1 peaks, as shown in Fig. 19 for comparison. Thus it is concluded that the origin of the 1540 cm-1 peaks is relevant to the nanotubes located within bundles. The interlayer interaction between layers of SWNTs is considered to be on the order of 5-50 cm-1 [29,48,49,50,52,53,54] and thus the difference between 1590 cm-1 and 1540 cm-1 is of about the same order of magnitude as the interlayer interaction. One open issue awaiting solution is why the 1540 cm-1 peaks are observed only when the metallic nanotube is within a bundle, and when the laser excitation is within the metallic window and corresponds to an interband transition contributing to the optical absorption. Thus the mechanism responsible for the 1540 cm-1 peak is not understood from a fundamental standpoint.

0 0

Post a comment