Page 267 - Adsorbents fundamentals and applications
P. 267

252   CARBON NANOTUBES, PILLARED CLAYS, AND POLYMERIC RESINS

                       The interesting idea of separating enantiomers via chiral separation with car-
                     bon nanotubes was suggested and examined by Sholl and co-workers (Power
                     et al., 2002). A Monte Carlo technique was used to calculate the isosteric heats
                     of adsorption for enantiomers of trans-1,2-dimethylcyclopropane (DMCP) and
                     trans-1,2-dimethylcyclohexane (DMCH) inside SWNTs. Sholl and colleagues
                     used tubes of different diameters as well as different chiral angles. Tube diameters
                                                                         ◦
                     ranging from 9.5 ˚ A to 28.7 ˚ A and chiral angles from 34 to 54 were examined.
                     Unfortunately, the isosteric heats of adsorption for the two pairs of enantiomers
                     were negligible in all cases.


                     Kinetic Separations. As discussed in Chapter 5, carbon molecular sieves have
                     already been used for gas separation that is based on differences in diffusivities of
                     different gas molecules. The same separations should also be possible with carbon
                     nanotubes. To this end, a number of simulation studies have been carried out.
                     Mao and Sinnott (2000 and 2001) have reported molecular dynamics simulation
                     results for diffusion of methane, ethane, n-butane, and isobutene, as well as their
                     binary mixtures, in SWNTs and their bundles. As expected, diffusion of smaller
                     molecules is faster, for example a factor of 25 was obtained for methane/isobutene
                     in a (8,8) nanotube (Mao and Sinnott, 2001).
                       The difficult separation of N 2 /CH 4 was studied with a SWNT of 13 ˚ Adiam-
                     eter by Nicholson and Suh (2002) by using a Monte Carlo technique. The flux
                     was expressed in the Fickian form to include both main-term and cross-term
                     diffusivities, as well as a viscous contribution. Their results are summarized in
                     Figure 9.14. In Figure 9.14, J 1 /J 2 is the ratio of the total fluxes of CH 4 over
                     N 2 , which reflects the overall separation. D 11 and D 22 are the main-term Fickian
                     diffusivities for CH 4 and N 2 , respectively. The large deviation of D 11 /D 22 from
                     J 1 /J 2 reflects the significant contribution of the cross-term diffusivities.




                                      16
                                             D /D 22
                                              11
                                    Selectivity  12 8          J /J 2
                                                                1



                                                   S (equilibrium)
                                       4
                                         0         2          4          6
                                                       r/nm −3
                     Figure 9.14. Comparison of equilibrium and kinetic selectivities of CH 4 over N 2 in a SWNT of
                                   ◦
                     13 ˚ Adiameter at25 C. The fugacities of CH 4 and N 2 are equal and ρ is the total adsorbate
                     density (Nicholson and Suh, 2002, with permission).
   262   263   264   265   266   267   268   269   270   271   272