Combining experimental knowledge with molecular simulations, we investigated the adsorption and separation properties of double-walled carbon nanotubes (DWNTs) against flue/synthetic gas mixture components (e.g. CO(2), CO, N(2), H(2), O(2), and CH(4)) at 300 K. Except molecular H(2), all studied nonpolar adsorbates assemble into single-file chain structures inside DWNTs at operating pressures below 1 MPa. Molecular wires of adsorbed molecules are stabilized by the strong solid-fluid potential generated from the cylindrical carbon walls. CO(2) assembly is formed at very low operating pressures in comparison to all other studied nonpolar adsorbates. The adsorption lock-and-key mechanism results from perfect fitting of rod-shaped CO(2) molecules into the cylindrical carbon pores. The enthalpy of CO(2) adsorption in DWNTs is very high and reaches 50 kJ mol(-1) at 300 K and low pore concentrations. In contrast, adsorption enthalpy at zero coverage is significantly lower for all other studied nonpolar adsorbates, for instance: 35 kJ mol(-1) for CH(4), and 14 kJ mol(-1) for H(2). Applying the ideal adsorption solution theory, we predicted that the internal pores of DWNTs have unusual ability to differentiate CO(2) molecules from other flue/synthetic gas mixture components (e.g. CO, N(2), H(2), O(2), and CH(4)) at ambient operating conditions. Computed equilibrium selectivity for equimolar CO(2)-X binary mixtures (where X: CO, N(2), H(2), O(2), and CH(4)) is very high at low mixture pressures. With an increase in binary mixture pressure, we predicted a decrease in equilibrium separation factor because of the competitive adsorption of the X binary mixture component. We showed that at 300 K and equimolar mixture pressures up to 1 MPa, the CO(2)-X equilibrium separation factor is higher than 10 for all studied binary mixtures, indicating strong preference for CO(2) adsorption. The overall selective properties of DWNTs seem to be superior, which may be beneficial for potential industrial applications of these novel carbon nanostructures.
Read full abstract