Technical Comments

Comment on "Tubular Graphite Cones"

See allHide authors and affiliations

Science  06 Feb 2004:
Vol. 303, Issue 5659, pp. 766
DOI: 10.1126/science.1090375

Zhang et al. (1) reported the formation of “tubular graphite cones” (TGCs), which may constitute a new addition to the family of carbon nanotube (CNT) materials. They characterized these cones and concluded that the graphite layers have identical zigzag-type chirality throughout the material. We disagree with their analyses and find that the chiralities of the constituent graphite sheets are not uniformly zigzag.

Zhang et al. presented two types of evidence to support their claim. First, they presented high-resolution transmission electron microscopy (HRTEM) images of TCG walls, which showed a dotted pattern. The authors compared the adjacent dot spacing (d1) resulting from projection of the sidewall region along the electron beam, with that expected for either zigzag or armchair chirality. They found that the observed spacing coincides with the projected spacing for zigzag chirality and concluded that the observed offset between dots in adjacent dot layers (d1/2) corresponds to the alternate layers of ABAB.. .graphite stacking. There are two problems with this analysis. The first is that the offset arising from the A and B layers, when projected in the sidewall region, is actually d1/3—not d1/2, as illustrated in Fig. 1 [same as figure 4A in (1) except that the atoms of the A and B layers are clearly distinguished]. The second, more serious issue concerns the interpretation of the HRTEM images. Because these images are generated from the interference of diffraction waves, their interpretation is not trivial. Figure 2A shows a fast Fourier transform (FFT) analysis of the HRTEM image showing dotted walls [figure 3D in (1)], in which the FFT pattern corresponds to two sets of diffraction spots: {101̄1} and {0002}. The HRTEM image shown in figure 3D in (1) cannot therefore be interpreted by a simple atomic projection, and can only be explained by considering the interference of the two sets of diffraction waves. The spacing (d1) and the offset (d1/2) in the sidewall HRTEM image are merely artifacts resulting from the interference of the two sets of diffraction waves. In contrast, a similar analysis of the HRTEM image from the core region [figure 3E in (1)] suggests that all the layers in this region indeed have identical chirality close to the zigzag type (Fig. 2B). The FFT pattern in Fig. 2B shows only six {101̄0} spots, distributed in a manner close to that of the six zigzag {101̄0} spots.

Fig. 1.

(A) Schematic representation of the cross section of a TGC or multi-wall CNT. The multiple cylindrical walls are simplified as a single circle. Planes X and Yare tangential to the circle. Slices of graphite sheets on the X plane can be deemed parallel to the electron beam, and those on the Yplane, normal to the electron beam. (B) A schematic image showing the ABAB.. .graphite layers in plane Y. Atoms of layer A are marked by small solid circles and atoms of layer B are marked by large open circles. The projections of the sidewall counterparts of these atoms along the electron beam (situated along plane X) are shown by large solid circles on the left.

Fig. 2.

(A) Analysis of figure 3D in (1). The inset is an FFT pattern obtained by superimposing 11 FFT images taken from 11 boxed HRTEM regions. (B) Analysis of figure 3E in (1). The inset is an FFT pattern obtained by superimposing 6 FFT images taken from 6 boxed HRTEM regions. (C) Analysis of figure 3F in (1). Apices of the hexagon mark the positions of spots expected for a zigzag tube. (D) SAED pattern taken from a helical multi-wall CNT. (E) HRTEM image of a sidewall region of the multi-wall CNT. The inset is an FFT pattern taken from the boxed region.

The second type of evidence presented by Zhang et al. in (1) was selected area electron diffraction (SAED) data taken from a different region of the TGC than the HRTEM images [figure 3F in (1)]. Analysis of the diffraction pattern using well-established methods for single- and multi-wall CNTs (26), whose graphite layers have a cylindrical and coaxial arrangement similar to TGCs, shows at least two helical chiralities (with angles of 8.6° and 18.3°; Fig. 2C). The diffraction spots corresponding to the helical chiralities are of significantly stronger intensity than the spots corresponding to the zigzag chirality. Hence, the most reasonable interpretation of this SAED pattern is that the chiralities in the corresponding region are predominately helical.

Figures 2 D and E provide additional evidence in support of the above arguments. These images were taken from a multi-wall CNT produced by arc discharge. The SAED pattern shown in Fig. 2D indicates that the constituent layers of the material possess many different helical angles (7). Figure 2E shows a HRTEM image from the sidewall region, where a dotted structure similar to that in (1) can be seen. As before, the appearance of this image is an artifact resulting from the interference of the {101̄1} and {0002} diffraction waves (as indicated by the FFT pattern in Fig. 2E). Thus, HRTEM images with dotted walls shown in figure 3D in (1) are not uniquely associated with identical zigzag chirality, but rather arise from interference of two sets of diffraction waves. Conclusions about chirality throughout the entirety of each TGC should therefore not be made based on the interpretation of these images alone.

References and Notes

Navigate This Article