Water under extreme confinement in graphene: Oscillatory dynamics, structure, and hydration pressure explained as a function of the confinement width

Graphene nanochannels are relevant for their possible applications, as in water purification, and for the challenge of understanding how they change the properties of confined liquids. Here, we use all-atom molecular dynamics simulations to investigate water confined in an open graphene slit-pore as a function of its width w, down to sub-nm scale. We find that the water translational and rotational dynamics exhibits an oscillatory dependence on w, due to water layering. The oscillations in dynamics correlate with those in hydration pressure, which can be negative (hydrophobic attraction), or as high as similar to 1 GPa, as seen in the experiments. At pore widths commensurable with full layers (around 7.0 angstrom and 9.5 angstrom for one and two layers, respectively), the free energy of the system has minima, and the hydration pressure vanishes. These are the separations at which the dynamics of confined water slows down. Nevertheless, the hydration pressure vanishes also where the free energy has maxima, i.e., for those pore-widths which are incommensurable with the formation of well-separated layers, as w similar or equal to 8.0 angstrom. Around these values of w, the dynamics is faster than in bulk, with water squeezed out from the pore. This behavior has not been observed for simple liquids under confinement, either for water in closed nano-pores. The decomposition of the free energy clarifies the origins of the dynamics speedups and slowdowns. In particular, we find that the nature of the slowdown depends on the number of water layers: for two layers, it is due to the internal energy contribution, as in simple liquids, while for one layer, it has an entropic origin possibly due to the existence of a hydrogen-bond network in water. Our results shed light on the mechanisms ruling the dynamics and thermodynamics of confined water and are a guide for future experiments. (C) 2020 Elsevier B.V. All rights reserved.