Pressure-Induced Stability of Methane Hydrate from Machine Learning Force Field Simulations

The stability of methane hydrate is essential for its practical applications as an energy resource, but the stability mechanism of methane hydrate under pressure has not been clarified at the molecular level. Here, we report the effect of pressure on the stability and decomposition of methane hydrate from molecular dynamics simulations using a quantum mechanics-derived machine-learning force field. We find that the decomposition behavior of methane hydrate is dominated by regulating the stability of the water cages. The water molecules have the largest stable amplitude when the distance of water molecules in cages is reduced at a 2 kbar pressure in which the distance is close to that in ice. Therefore, the most stable methane hydrate occurs at 2 kbar. In contrast, too low or too high pressure causes the distance of water molecules to deviate from the optimal value, which leads to the rapid deconstruction of the water cages and promotes the decomposition of methane hydrates. These findings provide molecular insight into the pressure-dependent decomposition of methane hydrates and may be extended to the stability of other gas hydrates under pressure.