Page 47 - Carbon Nanotubes
P. 47

38                            J. W. MINTMIRE and C. T. WHITE
                                                         sets, the nanotubes are chiral and have three inequiv-
                                                         alent helical operations.
                                                           Because real lattice vectors can be found that are
                                                         normal to the primitive  real  lattice  vectors  for  the
                                                         graphite sheet, each nanotube thus generated can be
                                                         shown to be translationally periodic down the nano-
                                                         tube axis[12-14,23].  However, even for relatively small
                                                         diameter nanotubes, the minimum number of atoms
                                                         in a translational  unit cell can be quite large. For ex-
                                                         ample, for the  [4,3] nanotube (nl = 4 and nz = 3)
                                                         then the radius of  the nanotube is less than 0.3 nm,
                                                         but the translational unit cell contains 148 carbon at-
                                                         oms as depicted  in Fig.  2.  The rapid  growth  in the
              Fig.  1.  Two-dimensional  honeycomb  lattice  of  graphene;
              primitive  lattice  vectors  R, and  R,  are depicted  outlining   number  of  atoms  that  can  occur  in  the  minimum
                            primitive unit cell.         translational unit cell makes recourse to the helical and
                                                         any higher point group symmetry of these nanotubes
                                                         practically mandatory in any comprehensive study of
              group symmetry) helical operations derived from the   their properties as a function of  radius and helicity.
              primitive lattice vectors of the graphite sheet. Thus,   These symmetries can be used to reduce  to two the
              while all nanotubes have a helical structure, nanotubes   number  of  atoms  necessary  to generate  any  nano-
              constructed by mapping directions equivalent to lat-  tube[13,14];  for example, reducing the matrices that
              tice translation indices of the form [n,O] and [n,n], to   have to be diagonalized in a calculation of the nano-
              the circumference  of the nanotube will possess a re-   tube’s electronic structure to a size no larger than that
              flection plane. These high-symmetry nanotubes will,   encountered  in a corresponding  electronic  structure
              therefore, be achiral[ 12-14,231.  For convenience, we   calculation of  two-dimensional  graphene.
              have  labeled  these  special  structures  based  on  the   Before we can analyze the electronic structure of
              shapes made by the most direct  continuous  path of   a nanotube in terms of its helical symmetry, we need
              bonds around the circumference of the nanotube[23].   to find an appropriate helical operator S (h, p) , rep-
              Specifically, the [n,O]-type structures were labeled as  resenting a screw operation with a translation h units
              sawtooth and the [n, n]-type structures as serpentine.   along the cylinder axis in conjunction with a rotation
              For all other conformations inequivalent to these two   p radians about this axis. We also wish to find the op-
                                                         erator S that requires the minimum unit cell size (Le.,
                                                         the smallest set of carbon atoms needed to generate
                                                         the entire nanotube using S) to minimize the compu-
                                                         tational complexity of calculating the electronic struc-
                                                         ture. We can find this helical operator S( h, p) by first
                                                         finding the real lattice vector H = mlR, + m2R2 in
                                                         the honeycomb lattice that will transform to S(h, p)
                                                         under conformal mapping, such that h = I H x BI /I  BI
                                                         and p = 2.1r(H.B)/(BIZ. An example of H is depicted
                                                         in Fig. 2 for the [4,3] nanotube. If we construct the
                                                         cross product H x B of H and B, the magnitude I H x
                                                         BI  will correspond to the area “swept out” by the he-
                                                         lical operator S. Expanding,  we  find



                                                         where I RI x Rz I is just the area per two-carbon unit
                                                         cell of  the primitive graphene lattice.  Thus (mln2 -
                                                         m2n1) gives the number of carbon pairs required for
                                                         the unit cell under helical symmetry with a helical op-
                                                         erator generated using H and the conformal mapping.
                                                         Given a choice  for n, and n2, then  (m~nz - mzn~)
                                                         must equal some integer value


              Fig. 2.  Depiction of conformal mapping of graphene lattice
              to [4,3] nanotube.  B denotes [4,3] lattice vector that trans-
              forms to circumference of nanotube, and H transforms into   Integer arithmetic then shows that the smallest mag-
              the helical operator yielding the minimum unit cell size un-   nitude nonzero value of N will be given by the great-
              der helical symmetry. The numerals indicate the ordering of
              the helical steps necessary to obtain one-dimensional trans-   est common divisor of nl and n2. Thus, we can easily
                            lation periodicity.          determine the helical operator H + S(h, p) that yields
   42   43   44   45   46   47   48   49   50   51   52