Calculating adsorption isotherms using Lennard Jones particle density distributions

Umbrella Sampling Simulations of Carbon Nanoparticles Crossing Immiscible Solvents

We use molecular dynamics to compute the free energy of carbon nanoparticles crossing a hydrophobic–hydrophilic interface. The simulations are performed on a biphasic system consisting of immiscible solvents (i.e., cyclohexane and water). We solvate a carbon nanoparticle into the cyclohexane layer and use a pull force to drive the nanoparticle into water, passing over the interface. Next, we accumulate a series of umbrella sampling simulations along the path of the nanoparticle and compute the solvation free energy with respect to the two solvents. We apply the method on three carbon nanoparticles (i.e., a carbon nanocone, a nanotube, and a graphene nanosheet). In addition, we record the water-accessible surface area of the nanoparticles during the umbrella simulations. Although we detect complete wetting of the external surface of the nanoparticles, the internal surface of the nanotube becomes partially wet, whereas that of the nanocone remains dry. This is due to the nanoconfinement of the particular nanoparticles, which shields the hydrophobic interactions encountered inside the pores. We show that cyclohexane molecules remain attached on the concave surface of the nanotube or the nanocone without being disturbed by the water molecules entering the cavity.

Pulling Simulations and Hydrogen Sorption Modelling on Carbon Nanotube Bundles

Recent progress in molecular simulation technology has developed an interest in modernizing the usual computational methods and approaches. For instance, most of the theoretical work on hydrogen adsorption on carbon nanotubes was conducted a decade ago. It should be insightful to reinvestigate the field and take advantage of code improvements and features implemented in contemporary software. One example of such features is the pulling simulation modules now available in many molecular dynamics programs. We conduct pulling simulations on pairs of carbon nanotubes and measure the inter-tube distance before they dissociate in water. We use this distance to set the interval size between adjacent nanotubes as we arrange them in bundle configurations. We consider bundles with triangular, intermediate and honeycomb patterns, and armchair nanotubes with a chiral index from n = 5 to n = 10. Then, we simulate low pressure hydrogen adsorption isotherms at 77 K, using the grand canonical Monte Carlo method. The different bundle configurations adsorb great hydrogen amounts that may exceed 2% wt at ambient pressures. The computed hydrogen capacities are considered large for physisorption on carbon nanostructures and attributed to the ultra-microporous network and extraordinary high surface area of the configured models.