Tuning Interfacial Water Friction through Moiré Twist.

Foundations of nanofluidics can enable advances in diverse applications such as water desalination, energy harvesting, and biological analysis. Dynamically manipulating nanofluidic properties, such as diffusion and friction, is an area of great scientific interest. Twisted bilayer graphene, particularly at the magic angle, has garnered attention for its unconventional superconductivity and correlated insulator behavior due to strong electronic correlations. The impact of the electronic properties of moiré patterns in twisted bilayer graphene on structural and dynamic properties of water remains largely unexplored. Computational challenges, stemming from simulating large unit cells using density functional theory, have hindered progress. This study addresses this gap by investigating water behavior on twisted bilayer graphene, employing a deep neural network potential (DP) model trained with a data set from ab initio molecular dynamics simulations. It is found that as the twisted angle approaches the magic angle, interfacial water friction increases, leading to a reduced water diffusion. Notably, the analysis shows that at smaller twisted angles with larger moiré patterns, water is more likely to reside in AA stacking regions than AB (or BA) stacking regions, a distinction that diminishes with smaller moiré patterns. This study illustrates the potential for leveraging the distinctive properties of moiré systems to effectively control and optimize interfacial fluid behavior.

Molecular transport enhancement in pure metallic carbon nanotube porins.

Nanofluidic channels impose extreme confinement on water and ions, giving rise to unusual transport phenomena strongly dependent on the interactions at the channel-wall interface. Yet how the electronic properties of the nanofluidic channels influence transport efficiency remains largely unexplored. Here we measure transport through the inner pores of sub-1 nm metallic and semiconducting carbon nanotube porins. We find that water and proton transport are enhanced in metallic nanotubes over semiconducting nanotubes, whereas ion transport is largely insensitive to the nanotube bandgap value. Molecular simulations using polarizable force fields highlight the contributions of the anisotropic polarizability tensor of the carbon nanotubes to the ion-nanotube interactions and the water friction coefficient. We also describe the origin of the proton transport enhancement in metallic nanotubes using deep neural network molecular dynamics simulations. These results emphasize the complex role of the electronic properties of nanofluidic channels in modulating transport under extreme nanoscale confinement.

From ab initio to continuum: Linking multiple scales using deep-learned forces.

We develop a deep learning-based algorithm, called DeepForce, to link ab initio physics with the continuum theory to predict concentration profiles of confined water. We show that the deep-learned forces can be used to predict the structural properties of water confined in a nanochannel with quantum scale accuracy by solving the continuum theory given by Nernst-Planck equation. The DeepForce model has an excellent predictive performance with a relative error less than 7.6% not only for confined water in small channel systems (L < 6 nm) but also for confined water in large channel systems (L = 20 nm) which are computationally inaccessible through the high accuracy ab initio molecular dynamics simulations. Finally, we note that classical Molecular dynamics simulations can be inaccurate in capturing the interfacial physics of water in confinement (L < 4.0 nm) when quantum scale physics are neglected.

Structural and Dynamical Properties of H2O and D2O under Confinement.

Water (H2O) is of great societal importance, and there has been a significant amount of research on its fundamental properties and related physical phenomena. Deuterium dioxide (D2O), known as heavy water, also draws much interest as an important medium for medical imaging, nuclear reactors, etc. Although many experimental studies on the fundamental properties of H2O and D2O have been conducted, they have been primarily limited to understanding the differences between H2O and D2O in the bulk state. In this paper, using path integral molecular dynamics simulations, the structural and dynamical properties of H2O and D2O in bulk and under nanoscale confinement in a (14,0) carbon nanotube are studied. We find that in bulk, structural properties such as bond angle and bond length of D2O are slightly smaller than those of H2O while D2O is slightly more structured than H2O. The dipole moment of D2O tends to be 4% higher than that of H2O, and the hydrogen bonding of D2O is also stronger than that of H2O. Under nanoscale confinement in a (14,0) carbon nanotube, H2O and D2O exhibit a smaller bond length and bond angle. The hydrogen bond number decreases, which demonstrates a weakened hydrogen bond interaction. Moreover, confinement results in a lower libration frequency and a higher OH(OD) bond stretching frequency with an almost unchanged HOH(DOD) bending frequency. The D2O-filled (14,0) carbon nanotube is found to have a smaller radial breathing mode than the H2O-filled (14,0) carbon nanotube.