Laboratoire Charles Coulomb UMR 5221 CNRS/UM2 (L2C)


Accueil > La Recherche > Axes & Equipes > Nanostructures & Spectroscopies > Equipe : Nanomatériaux > Thème : Propriétés intrinsèques des nanotubes individuels et du graphène

Vibrational properties of individual carbon nanotubes

par Sébastien LAYSSAC - publié le

Involved researchers : D. Nakabayashi, T. Michel, R. Parret, M. Paillet, J.-L. Sauvajol, A. Zahab
Collaborations : J. C. Meyer, S. Roth (MPI Stuttgart, Germany), V. Popov (University of Sofia, Bulgaria), L. Henrard (FUNDP, Belgium), Philippe Roussignol, C. Voisin (ENS Paris), A. Loiseau, R. Arenal (ONERA, Châtillon), A. San Miguel (LPMCN, Lyon), B. Lounis, L. Cognet (CPMOH, Bordeaux)

The resonant Raman scattering of light by carbon nanotubes is a powerful tool for the study of their physical properties and for their characterization. The physical properties and Raman signature of individual carbon nanotubes are related to their atomic structure (diameter and chirality). Raman studies at the single nanotube level are necessary because the large disparity of nanotube types present in bulk samples complicate data interpretation. Our research activities are devoted to determine the intrinsic Raman spectra of individual and structure identified carbon nanotubes. We have developed an approach combining Raman spectroscopy and electron diffraction experiments on individual, spatially isolated, freestanding single-walled carbon nanotubes.

The pioneer experiments developed in our group consist to combine Raman experiments with a precise atomic structure determination by electron diffraction of the same individual, spatially isolated, freestanding carbon nanotube. Before the Raman experiments, overview transmission electron microscopy (TEM) images are recorded in order to determine the position and orientation of nanotubes (Step 1). Raman spectra are measured on those located nanotubes at fixed excitation energy with the incident light polarized along the tube axis (step 2). If the excitation laser energy matches an optical transition of the tube under investigation, the measurement of a Raman signal can be done.The structure identification of the tubes for which a Raman signal was measured is obtained from electron diffraction (step 3). The experimental diffraction pattern is compared with a simulation giving unambiguously the atomic structure of the nanotubes.

Figure 1 : Experimental procedure to combine Raman spectroscopy and electron diffraction on the same individual, spatially isolated, freestanding nanotube - © L2C

We investigated the two main first-order Raman modes of carbon nanotubes : the Radial Breathing Mode (RBM) and the G-modes. Figure 2 shows the RBM and G-modes measured on three chiral semiconducting tubes : (11, 10), (17, 9) and (27, 4). The RBM is specific to carbon nanotubes, its frequency is inversely proportional to the diameter. As expected for an individual nanotube, a single RBM is observed in each Raman spectrum. Figure 3 shows the first experimental dependence of the RBM frequencies with the diameter obtained from the diffraction pattern of individual nanotubes.
In the G-mode region, two components are observed for chiral nanotubes. In agreement with modeling, each component is assigned to a A symmetry mode. The highest-frequency mode (assigned to the LO mode in semiconducting nanotubes and to the TO mode in metallic nanotubes) does not significantly change with the diameter of the tubes. By contrast, the lowest-frequency mode (assigned to the TO mode in semiconducting nanotubes and the LO mode in metallic nanotubes) significantly depends on diameter and chirality. For semiconducting tubes, the TO modes frequencies are in good agreement with phonon calculations that include the curvature effects. For metallic nanotubes, our results support the non adiabatic dynamic of the phonon modes. Our results evidenced the single resonance process as the dominant process for first-order Raman scattering in SWNTs.

Figure 2 : (a) Radial Breathing Mode and (b) G-modes of semiconducting carbon nanotubes for which the atomic structures are determined by electron diffraction - © L2C

Figure 3 : Left : first experimental dependence of the RBM frequencies with the diameter. Right : semi-experimental resonance chart for the optical transitions in carbon nanotubes. Experiments : red symbols, calculations : black and blue symbols - © L2C

In a resonant process the detection of a Raman signal in the RBM range in a short counting time means that the laser excitation energy is very close to an optical transition of the nanotube. The measurement of a Raman spectrum then provides an estimation of the transition energy. To analyze our results we compared the laser energies used for Raman measurements with calculated values of the optical transition using a non-orthogonal tight-binding approach (NTB). We found that the first optical transition of metallic tubes and the third and fourth transitions of semiconducting tubes have to be corrected by non-diameter-dependent shifts of 0.3 eV and 0.4 eV, respectively. The normalization of the NTB calculations provides a precise resonance chart of the nanotubes in a broad diameter/energy range (figure 3).

From our experimental results on the phonon frequencies and the optical energies we have been able to define Raman criteria that allow the precise atomic identification of individual nanotubes from Raman spectroscopy only.
The diameter of the nanotube is derived from the measurement of its RBM frequency. The identification is achieved by reporting the couple (diameter, laser excitation energy) in the normalized resonance chart. The matching of the experimental coordinates with a calculated nanotube in the resonance chart gives the atomic structure of the nanotube under investigation. The G-mode lineshape and frequencies must be in agreement with the atomic structure to ensure that the Raman spectrum indeed corresponds to a nanotube that is not part of a bundle or of a small multiwall nanotube. By this way, we established the intrinsic Raman responses for all possible chiralities of semiconducting and metallic carbon nanotubes : chiral, zigzag and armchair (figure 4).

Figure 4 : Intrinsic tangential modes for all possible chiralities of SWCNTs : zigzag and chiral semiconducting (25,0) and (11,10) SWCNTs and metallic chiral, zigzag, armchair (19,16), (24,0), (10,10) SWCNTs - © L2C

M. Paillet, Thèse de l’Université Montpellier 2 (2005), M. Paillet et al., PRL 94, 237401 (2005), J.C. Meyer et al., PRL 95, 217401 (2005), M. Paillet et al., PRL 96, 257401 (2006), T. Michel et al., PRB 75, 155432 (2007), T. Michel, Thèse de l’Université Montpellier 2 (2007), T. Michel et al., PRB in press (2009).

Fundings : ANR PNANO Nanotubes suspendus and T-NICE, ANR P3N Excitubes.