Adsorption and separation of linear and branched alkanes on carbon nanotube bundles from configurational-bias Monte Carlo simulation

Abstract

The adsorption and separation of linear ͑C 1 -nC 5 ͒ and branched ͑C 5 isomers͒ alkanes on single-walled carbon nanotube bundles at 300 K have been studied using configurational-bias Monte Carlo simulation. For pure linear alkanes, the limiting adsorption properties at zero coverage exhibit a linear relation with the alkane carbon number; the long alkane is more adsorbed at low pressures, but the reverse is found for the short alkane at high pressures. For pure branched alkanes, the linear isomer adsorbs to a greater extent than its branched counterpart. For a five-component mixture of C 1 -nC 5 linear alkanes, the long alkane adsorption first increases and then decreases with increasing pressure, but the short alkane adsorption continues increasing and progressively replaces the long alkane at high pressures due to the size entropy effect. All the linear alkanes adsorb into the internal annular sites with preferred alignment parallel to the nanotube axis on a bundle with a gap of 3.2 Å, and also intercalate the interstitial channels in a bundle with a gap of 4.2 Å. For a three-component mixture of C 5 isomers, the adsorption of each isomer increases with increasing pressure until saturation, though nC 5 increases more rapidly with pressure and is preferentially adsorbed due to the configurational entropy effect. All the C 5 isomers adsorb into the internal annular sites on a bundle with a gap of 3.2 Å, but only nC 5 also intercalates the interstitial channels on a bundle with a gap of 4.2 Å. This work suggests the possibility of separating alkane mixtures based on differences in either size or configuration, as a consequence of competitive adsorption on the carbon nanotube bundles.

Adsorption of alkanes, alkenes and their mixtures in single-walled carbon nanotubes and bundles

“…Note that both, the chain length dependency of the Henry coefficients and the chain length dependency of the heats of adsorption were also observed experimentally for alkanes in zeolites [40]. Furthermore, these dependencies were already observed with molecular simulation methods for the adsorption of alkanes in CNT bundles [4] and, in agreement with experiments, for the adsorption in silicalite [20,41].…”

“…These maxima were also observed by Jiang et al [4] and Heyden et al [22] for carbon nanotubes and by Jakobtorweihen et al [24] for zeolites, among others. Note that Jiang et al [4] have also observed distinct maxima in the selectivities for mixtures of C5 isomers. The selectivity drops from about two at low pressures to about one at high pressures.…”

“…Besides hydrocarbons [2][3][4][5][6], adsorption of other molecules and mixtures on single-walled carbon nanotubes (SWCNTs) were investigated, like hydrogen [7][8][9], nitrogen [10][11][12][13], nitrogen/oxygen [14,15], water [16], neon [17], helium [18], and xenon [19]. Cao et al [3] investigated the adsorption of methane on triangular arrays of SWCNTs at room temperature with the grand canonical Monte Carlo (GCMC)…”

“…The gap between the tubes was varied in order to find the optimum gap with respect to the amount of adsorbed methane. Jiang et al [4] studied adsorption and separation of linear (C1-C5) and branched (C5 isomers) alkanes on (10,10) SWCNT bundles at 300 K using configurationalbias Monte Carlo (CBMC) simulations. For pure linear alkanes, the limiting adsorption properties at zero coverage exhibit a linear relation with the alkane carbon number; long alkanes are more adsorbed at low pressures, but the reverse is found for short alkanes at high pressures.…”

Diffusion of chain molecules and mixtures in carbon nanotubes: The effect of host lattice flexibility and theory of diffusion in the Knudsen regime

“…Thus far, only a few studies deal with the behavior of alkanes in nanotubes. [34][35][36][37][38][39][40][41][42][43][44] Further, Wei and Srivastava 20 investigated the behavior of polymers in CNTs.…”