Transmission Electron Microscopy and Electron Diffraction

The existence of SWNTs was first confirmed by means of high-resolution electron microscopy and electron diffraction. Due to the small physical size sample and the low atomic number of carbon, classical structural techniques such as X-ray diffraction are difficult to apply to SWNTs. For this reason, TEM has been the most widespread tool for structural characterization of SWNTs.

Figure 2 shows a typical TEM micrograph of a SWNT rope. One of the first observations obtained from TEM was that SWNTs have a tendency to assemble in bundles containing between 10 and 50 aligned nanotubes. It is generally accepted that the SWNTs are held together by van der Waals interactions. Another interesting feature is also observed in Figure 2. A cross-sectional view shows that the individual SWNTs can be clearly distinguished. This is a common observation during TEM analysis and is due to the fact that the SWNT bundles are usually bent and it is possible that a portion of them get oriented parallel to the electron beam, resulting in images of the nanotube ends. TEM images show that the SWNT bundles are typically 5-20 nm in diameter and exhibit a triangular lattice with an inner nanotube distance of around 1.7 nm [15, 39].

An important precaution to be taken into consideration when performing TEM measurements on SWNT samples is that they are highly susceptible to the 100-200 keV electron beam of the TEM instrument. Beam damage makes it very difficult to carry out electron diffraction studies from isolated SWNTs. Diffractions patterns can be more readily obtained from bundles of SWNTs. By using low electron beam currents it has been possible to obtain good TEM measurements of lattice fringe images and electron diffraction patterns on single-wall nanotubes. These studies have shown that the walls of SWNTs have indeed the local structure of a graphene sheet [3].

The diffraction patterns to be expected from a SWNT sample can be approximated by forming a reciprocal space distribution for a single graphene sheet and rotating it around the nanotube axis to give the scattering power in a set of planes [40]. The interpretation of these electron diffraction patterns becomes easier by the use of computer simulations, which yield the expected diffraction pattern for an individual SWNT with well-defined (n, m) indexes [41-43].

For instance Figure 3a shows the diffraction pattern calculated for a bundle tilted by an angle of 60° with respect to the electron beam [44]. In this simulation, the bundle

Figure 2. High resolution TEM images of SWNT bundles produced by arc discharge technique. (a) Cross-section-like view of a polycrystalline bundle. (b) View of a twisted bundle with its axis normal to the electron beam. The inset shows a magnication of fringes related to (11) and (10) lattice planes. Reprinted with permission from [44], L. Henrard et al., Eur. Phys. J. B 13, 661 (2000). © 2000, Springer-Verlag.

Figure 2. High resolution TEM images of SWNT bundles produced by arc discharge technique. (a) Cross-section-like view of a polycrystalline bundle. (b) View of a twisted bundle with its axis normal to the electron beam. The inset shows a magnication of fringes related to (11) and (10) lattice planes. Reprinted with permission from [44], L. Henrard et al., Eur. Phys. J. B 13, 661 (2000). © 2000, Springer-Verlag.

is assumed to be composed of 55 nanotubes corresponding to eight different helicities: (10, 10), (11, 9), (12, 8), (13, 7), (14, 5), (15, 4), (16, 2), and (17, 0). Two rings of different intensity can be distinguished in the simulated diffraction pattern. Figure 3b shows an electron diffraction pattern experimentally obtained on an isolated straight bundle of SWNTs. The simulation reproduces very nicely the experimental pattern. Since the experimental data could not be simulated with a single chirality, the authors proposed that that SWNT bundles have to be composed of SWNTs with random chirality. The observed patterns need to be compared to the results of the modeling since the computer simulations are not able to carry out the inverse process of

Figure 3. Left: A simulated electron diffraction pattern obtained considering a single bundle composed of 55 nanotubes corresponding to eight different helicities: (10,10), (11, 9), (12, 8), (13, 7), (14, 5), (15, 4), (16, 2), and (17, 0). Right: Experimental diffraction pattern of an isolated straight SWNT bundle. Reprinted with permission from [44], L. Henrard et al., P. Bernier, Eur. Phys. J. B 13, 661 (2000). © 2000, Springer-Verlag.

Figure 3. Left: A simulated electron diffraction pattern obtained considering a single bundle composed of 55 nanotubes corresponding to eight different helicities: (10,10), (11, 9), (12, 8), (13, 7), (14, 5), (15, 4), (16, 2), and (17, 0). Right: Experimental diffraction pattern of an isolated straight SWNT bundle. Reprinted with permission from [44], L. Henrard et al., P. Bernier, Eur. Phys. J. B 13, 661 (2000). © 2000, Springer-Verlag.

converting an observed diffraction pattern into a real space structure (n, m) for the atomic sites within the 1D unit cell.

0 0

Post a comment