In Silico Engineering of Hydrate Anti-agglomerant Molecules Using Bias-Exchange Metadynamics Simulations

We use bias-exchange metadynamics simulations to quantify the adsorption of anti-agglomerant molecules to the surface of the sII methane-propane gas hydrate. Multiple steps of sampling binding configurations, measuring their free energies, and finding their equilibrium distribution are combined into a single-step modeling using bias-exchange metadynamics. We showed that the current method quantitatively and qualitatively reproduces the previously measured binding free energies for the adsorption of the quaternary ammonium salt to the hydrate surface. Finally, we used the developed approach together with a rational chemical design to engineer the tail of the AA molecule. We used the quaternary ammonium salt, which has been proven to be effective AA in experimental conditions as the initial configuration. This AA has a long flexible tail, which causes a sizeable entropic penalty upon binding, which could be reduced by adding a rigid section to it. We tested multiple ring structures with a higher rigidity compared to the alkyl tail and modified them to have new properties. We showed that among examined ring structures, the fluorene ring with an oxygen atom makes the most considerable decrease in standard binding free energy compared to the original molecule, which is 0.9 kcal/mol. We believe that the reason for the improved performance is threefold. First, a rigid tail decreases the entropic penalty upon binding. Second, the oxygen atom has a more favorable interaction with the hydrate crystal compared to the methylene group. Third, the proper size of the ring gives the AA molecule the chance to bind to the hydrate surface with its head, tail, and the ring structure.