Page 263 - Adsorbents fundamentals and applications

248 CARBON NANOTUBES, PILLARED CLAYS, AND POLYMERIC RESINS

The high selectivity for NO x and SO 2 over CO 2 simply reflects the fact that the
surfaces of the nanotubes are the basal plane of graphite. Subsequent measure-
ments by Long and Yang (unpublished results) with activated carbon showed
that the amounts adsorbed of NO x and SO 2 were similar tothose byMWNTs;
however, the amount of CO 2 was also high, that is, activated carbon did not have
the selectivity for NO x and SO 2 over CO 2 . Moreover, the amounts adsorbed of
NO x and SO 2 on MWNTs were significantly higher than those on activated car-
bon when normalized based on surface area, since the surface area of activated
carbon was more than twice that of the MWNTs. The enhanced adsorption by
MWNTs was most likely due to the cylindrical pore geometry.

Molecular Simulations and Adsorption of Other Gases. The first molecular
simulation study on adsorption in carbon nanotubes was reported by Pederson and
Broughton (1992). The potential curves were calculated for two HF molecules,
approaching from the two opposite ends of a SWNT. Two tubes were subjected to
calculations, with inner diameters of 8.2 and 10 ˚ A. At the equilibrium distance of
approximately 3 ˚ A between the two HF, collinearly positioned along the center
of the tube axis, a potential energy of approximately −0.3 eV was obtained.
This energy is equivalent to approximately 17.6 kJ/mol. More interestingly, they
3
2
arrived at an effective polarizability per unit area of the nanotube of 3.4 ˚ A / ˚ A .
This value may be compared with the polarizability of the carbon atom at the
3
ground state. Such a value is 1.76 ˚ A (see Table 2.2 in Chapter 2). Thus, the
inner tube surface appears to have an enhanced polarizability.
GCMC (grand canonical Monte Carlo) simulations of adsorption of N 2 on both
SWNTs (Maddox and Gubbins, 1995; Yin et al., 1999; Ohba et al., 2001; Mays
et al., 2002) and MWNTs have been performed. Experimental data are also avail-
able (Inoue et al., 1998; Ohba et al., 2001; Mays et al., 2002). Interesting yet not
unexpected results are seen. Maddox and Gubbins (1995) studied both adsorption
of argon and nitrogen at 77 K on three nanotubes with 1.02 and 4.78 nm diam-
eters. For the small tube, the isosteric heats of adsorption was approximately 16
and 17 kJ/mol for Ar and N 2 , respectively. The heats of adsorption decreased
to approximately 12 kJ/mol for the large tubes. For tubes shorter than 30 nm,
strong end effects were predicted.
Simulations as well as experiments have been performed on the adsorption of
4
He on bundles of SWNTs (Teizer et al., 1999; Teizer et al., 2000; Cole et al.,
2000; Talapatra et al., 2000; Gatica et al., 2000). From desorption measurements
at T> 14 K, a large binding energy of 330 K was first reported by Teizer et al.
(1999). This was 2.5X the value on planar graphite. They subsequently corrected
it to 1.6X (Teizer et al., 2000). Adsorption isotherms and binding energies of Xe,
CH 4 , and Ne on bundles of SWNTs were measured by Talapatra et al. (2000).
Again, higher binding energies than those on planar graphite were obtained.
Remarkably, the percent increase in the binding energy relative to planar graphite,
at about 75% for all three gases, was quantitatively the same. The temperatures
of their measurements were up to 296 K for Xe, 195 K for CH 4 , and 57 K
for Ne. The binding energies were 222 meV for CH 4 , 282 meV for Xe, and