Page 52 - Carbon Nanotubes
P. 52

Electronic and structural  properties of  carbon nanotubes   43
           lier workers to fit the exchange-correlation 'potential
           and the charge density (in the Coulomb potential) to
           a linear combination of  Gaussian-type functions.
             We have carried out a series of geometry optimi-  where E is the elastic modulus, R is the radius of cur-
           zations on nanotubes with diameters less than 2 nm.   vature, L is the length of the cylinder, and a is a rep-
           We will present some results for a selected subset of   resentative  thickness  of  the order  of  the  graphite
           the moderate band gap nanotubes, and then focus on  interplanar spacing (3.35 A). Assuming that the total
           results for an example chiral systems: the chiral [9,2]   number of carbons is given by N = 23rRL/hl, where
           nanotube with a diameter of 0.8 nm. This nanotube   hl is the area per carbon, we find that the strain energy
           has been chosen because its diameter corresponds  to   per carbon is expected to be
           those found in relatively large amounts by Iijima[7]
           after the synthesis of single-walled nanotubes.
             How  the structural  properties of  the  fullerene
           nanotubes  change with  conformation  is one of  the
           most important questions to be answered about these   In earlier work, we found this relationship was well
           new materials.  In particular, two properties are most  observed, using empirical bond-order potentials for all
           apt for study with the LDF approach:  how does the  nanotubes with radii less than 0.9  nm, and for a range
           band gap change with nanotube diameter, and how  of serpentine nanotubes using the LDF method. As
           does the strain energy change with nanotube diameter?  part of our studies, we have carried out first-principles
           We have studied these questions extensively using em-  LDF calculations on a representative sample of chi-
           pirical methods[ 10-14,23], and are currently working  ral nanotubes.  Iijima and Ichihashi[7] have recently
           on a comprehensive study of the band gaps, strain en-  reported the synthesis of  single-shell fullerene nano-
           ergies, and other properties using the LDF approach.  tubes with diameters of about 1 nm, using the gas-phase
           We expect from eqn (7) and the previous analysis of  product  of  a  carbon-arc  synthesis  with  iron  vapor
           the graphene sheet model that the moderate band gap  present.  After plotting the frequency of  single-shell
           nanotubes should have band gaps that vary roughly in-  nanotubes  they observed versus nanotube diameter,
           versely proportional to the nanotube radius. In Fig. 6  they  found enhanced  abundances  for diameters  of
           we  depict  representative  results  for some moderate  roughly 0.8 nm and 1.1 nm. For first-principles sim-
           band gap nanotubes.  In the figure,  we  see that not  ulations then to be useful, they should be capable of
           only do our first-principles band gaps decrease in an  calculations for nanotubes with diameters from about
           inverse relationship to the nanotube radius, but that  0.6-2.0  nm. In Fig. 7 we depict the calculated total en-
           the band gaps are well  described with a reasonable  ergy per carbon, shifted  relative to an extrapolated
           value of  V,, - -2.5  eV.                  value for an infinite radius nanotube, for a represen-
             Second, we expect that the strain energy per car-  tative sample of nanotubes over this range using the
           bon  should  increase  inversely  proportional to  the  LDF approach. In this figure, the open squares denote
           square of the nanotube radius[23]. Based on a contin-  results for unoptimized  nanotubes,  where the nano-
           uum elastic model, Tibbetts[4] derived a strain energy
           for a thin graphitic nanotube of the general form:

                                                          0.4



                                                          0.3




                                                        c
                                                        2
                                                       .A
                                                       4
                                                          0.1


                                                           0
                                                                   Nanotube  radius  (nm)
                0
                 0         0.5        1         1.5   Fig. 7.  Strain energy per  carbon  (total  energy minus total
                        Nanotube  radius  (nm)        energy extrapolated for the graphite sheet) as a function of
                                                      nanotube radius calculated for unoptimized nanotube struc-
           Fig. 6. Band gap as a function of nanotube radius using first-   tures (open squares) and optimized nanotube structures (solid
           principles  LDF method.  Solid line  shows estimates  using   circles). Solid line depicts inverse square relationship drawn
                graphene sheet model with  V,,   = -2.5  eV.   through point at smallest radius.
   47   48   49   50   51   52   53   54   55   56   57