Molecular simulations of gas adsorption on single walled carbon nanotube bundles composed of tubes of all the same diameter (homogeneous) and tubes of different diameters (heterogeneous) have been performed and compared with experiments [1] in an attempt to develop a nanotube bundle model that best explains the experimental results. Adsorption isotherms, isosteric heats of adsorption, and specific surface areas have been computed for Ar, CH₄, and Xe on closed, open, and partially opened homogeneous and heterogeneous bundles. These same quantities have previously been measured for these gases on unpurified HiPco nanotubes [1]. Closed homogeneous bundles give adsorption isotherms that are an order of magnitude smaller than experimental data at low pressures and give isosteric heats that are inconsistent with experimental results. In contrast, closed heterogeneous bundles give isotherms in qualitative and in some cases quantitative agreement with experiments at low coverage. An open homogeneous bundle consisting of (8,8) tubes also exhibits isosteric heats that are close to the experimental values at low coverage for all three gases, but fails to predict isosteric heats at high coverage and predicts adsorption isotherms that are qualitatively different from experiments. Homogeneous and hetergeneous bundles with approximately 10% open tubes have been studied for Xe at low to high coverages. These simulations yield isosteric heat values that are consistent with experiments over a wide range of coverages, suggesting bundles with a fraction of open tubes may serve as useful models for bundles used in experiments. Specific surface areas can be estimated from experiments and simulations through the point-B method applied to measured isotherms. Specific surfaces areas for closed, open and fractionally open homogeneous and heterogeneous bundles have been calculated for all three gases. The trends computed from partially opened heterogeneous bundles are in the best agreement with the experiments. This indicates that the partially opened heterogeneous bundle is the best overall model for HiPco nanotubes.

[1] V. Krungleviciute, L. Heroux, S. Talapatra and A. D. Migone, Nano Lett., 4, 1133-1137 (2004).