From ab initio intermolecular potentials to predictions of macroscopic thermodynamic properties and the global distribution of gas hydrates

Currently, the modeling of fluid and solid thermophysical properties depends strongly on accurate and numerous experimental measurements. Since this can be quite time consuming and expensive, especially for explosive or high-pressure systems, ab initio quantum mechanics (QM) offers an alternative to predict properties without experiments. Gases adsorbing in solid water hydrates and carbon structures were studied, which require accurate measures of gas-solid interaction energies. Previously, potential model parameters are fit to experiments, but in some instances, these parameters result in the inability to predict unknown phase behavior or the true gas-solid potential.; Gas hydrates, a solid network of water molecules with cavities that encapsulate guests, can exist in oil pipelines, permafrost regions, and ocean sediment. Thermodynamic models for gas hydrates require intermolecular potentials to predict cage occupancies by gas molecules. Instead of empirically fitting these potentials, a method to obtain accurate gas-water potentials from QM is presented. These QM-based potentials result in accurate predictions of cage occupancies for single and mixed gas hydrates, when compared to NMR and Raman spectroscopy measurements. A fugacity-based model using these ab initio potentials was developed and found to greatly improve equilibrium pressure predictions compared with a previous model. Finally, the global distribution of marine methane hydrate was estimated to be three orders more methane in hydrate form compared with conventional natural gas reserves.; Molecular simulation can be used to better understand the diffusion and adsorption behavior of gas molecules in nanoporous carbon (NPC) membranes. However, accurate gas-carbon force fields are required, where limited or no experimental data is available. The interactions between N₂(O ₂) and C₁₆₈ carbon schwarzite (a hypothetical carbon structure similar to experimental NPC membranes with pores and a curved carbon surface) were calculated from QM and compared to graphite-based potentials. The QM-based potential well depths for N₂-C168 interactions were found to be about 40% more attractive than those developed from graphite-based potentials, while O₂-C₁₆₈ interactions were 5% less attractive than the graphite-based potentials. Therefore, the common practice of using gas-graphite intermolecular potentials to determine the separation factor of nitrogen and oxygen in NPCs may be invalid.