Page 17 - Carbon Nanotubes
P. 17

Pyrolytic carbon nanotubes from vapor-grown carbon  fibers
              The toroidal  structures show interesting  changes
            in morphology as they become larger-at  least at the
            lip. The hypothetical small toroidal structure shown
            in Fig. 1 lb is actually quite smooth and has an essen-
            tially rounded structure[24]. As the structures become
            larger, the strain tends to focus in the regions near the
            pentagons  and heptagons,  and this  results  in  more
            prominent localized cusps and saddle points. Rather
            elegant toroidal structures with Dnh and Dnd symme-
            try are produced,  depending on whether the various
            paired heptagodpentagon sets which lie at opposite
            ends of the tube are aligned or are offset. In general,
            they probably lie is fairly randomly disposed positions.
            Chiral structures can be produced by off-setting the
            pentagons and heptagons.  In the D5d structure shown
            in Fig. 11 which was developed for the basic study, the
            walls  are fluted  between the heptagons  at opposite
            ends of the inner tube and the pentagons of the outer
            wall rim[l7]. It is interesting to note that in the com-
            puter images the localized cusping leads to variations
            in the smoothness of the image generated by the rim,
            though  it  still  appears  to  be  quite  elliptical  when
            viewed at an angle[ 171. The observed image appears
            to exhibit variations that are consistent with the local-     c  B
            ized cusps as the model predicts.                                 k
              In this study, we note that epitaxial graphitization   Spinale-shape model
            is achieved by heat treatment of the apparently mainly
            amorphous material which surrounds a single-walled   Fig.  12.  As-grown PCNTs with partially thickened spindle
                                                       shape (a) and the proposed structural model for spindle par-
            nanotube[ 171. As well as bulk graphitization, localized   ticles including  12 pentagons in hexagon cage (b).
            hemi-toroidal structures that connect adjacent walls
            have been identified and appear to be fairly common
            in this type of  material.  This type of  infrastructure  ber  of  pentagons  as required  by  variants of  Euler’s
            may be important as it suggests that double walls may  Law. Hypothetical structural models for these spin-
            form fairly readily.  Indeed, the observations suggest  dles are depicted  in Fig.  12. It is possible that simi-
            that pure carbon rim-sealed structures may be readily  lar  two-stage  growth processes occur  in the case of
            produced by heat treatment, suggesting that the future  ACNTs but, in general, the secondary growth appears
            fabrication  of  stabilized  double-walled  nanoscale   to be intrinsically  highly epitaxial.  This may be  be-
            graphite  tubes  in  which  dangling  bonds  have  been  cause in the ACNT growth case only carbon atoms are
            eliminated is a feasible objective. It will be interesting  involved and there are fewer (non-graphitizing) alter-
            to prove the relative reactivities of these structures for  native accretion pathways available. It is likely that
            their possible future applications in nanoscale devices  epitaxial  growth control  factors will be rather  weak
            (e.g.,  as quantum wire supports).  Although the cur-   when  secondary deposition is very fast, and so thin
            vatures of the rims appear to be quite tight, it is clear   layers may result in poorly ordered graphitic structure
            from the abundance of loop images observed, that the  in the thicker sections. It appears that graphitization
            occurrence of such turnovers between concentric cylin-  of this secondary deposit that occurs upon heat treat-
            ders with a gap spacing close to the standard graphite  ment may be partly responsible for the fine structure
            interlayer spacing is relatively common. Interestingly,  such  as compartmentalization,  as  well  as  basic  tip
            the edges of the toroidal structures appear to be readily  morphology[ 171.
            visible and this has allowed us to confirm the relation-
            ship between opposing loops. Bulges in the loops of
            the kind observed are simulated theoretically[ 171.   7.  VGCFs DERIVED FROM NANOTUBES
               Once one layer has formed (the primary nanotube   In Fig. 13 is shown the 002 lattice images of an “as-
            core), further secondary layers appear to deposit with   formed” very thin VGCF. The innermost core diam-
            various degrees of epitaxial coherence. When inhomo-  eter (ca. 20 nm as indicated by arrows) has two layers;
            geneous deposition occurs in PCNTs, the thickening   it  is  rather  straight  and appears to  be  the  primary
            has  a  characteristic  spindle  shape,  which may be  a   nanotube. The outer carbon layers, with diameters ca.
            consequence of non-carbon impurities which impede   3-4  nm, are quite uniformly stacked  parallel to the
            graphitization  (see below)-  this  is  not  the case  for  central core with 0.35 nm spacing. From the difference
            ACNTs were growth takes place in an essentially all-  in structure as well as the special features in the me-
            carbon atmosphere, except, of course, for the rare gas.   chanical strength (as in Fig. 7) it might appear possi-
            These spindles probably include the appropriate num-  ble that the two intrinsically different types of material
   12   13   14   15   16   17   18   19   20   21   22