Atomistic Hydrodynamics and the Dynamical Hydrophobic Effect in Porous Graphene

Abstract

Mirroring their role in electrical and optical physics, two-dimensional crystals are emerging as novel platforms for fluid separations and water desalination, which are hydrodynamic processes that occur in nanoscale environments. For numerical simulation to play a predictive and descriptive role, one must have theoretically sound methods that span orders of magnitude in physical scales, from the atomistic motions of particles inside the channels to the large-scale hydrodynamic gradients that drive transport. Here, we use constraint dynamics to derive a nonequilibrium molecular dynamics method for simulating steady-state mass flow of a fluid moving through the nanoscopic spaces of a porous solid. After validating our method on a model system, we use it to study the hydrophobic effect of water moving through pores of electrically doped single-layer graphene. The trend in permeability that we calculate does not follow the hydrophobicity of the membrane but is instead governed by a crossover between two competing molecular transport mechanisms.

Thermodynamic formalism for transport coefficients with an application to the shear modulus and shear viscosity

“…Here, we have studied the shear modulus and shear viscosity of a model system and found excellent agreement with other numerical methods. This thermodynamic formalism allows to calculate transport coefficients from fluctuations in a rigorous way without perturbing the dynamics, and without introducing constrained dynamics [14] or unphysical forces/moves (swapping momenta, displacing particles, . .…”

“…constant current). Simulating at constant stress might offer new physical insights [9][10][11], but requires some sort of feedback or "constrained simulations" [12][13][14]. While equilibrium properties are defined by the Boltzmann factor and thus energies, the situation is different for dynamics breaking detailed balance since the characteristics of the ensuing non-equilibrium steady state do depend on these dynamics.…”

Assessment of Density Functional Theory in Predicting Interaction Energies Between Water and Polycyclic Aromatic Hydrocarbons: From Water on Benzene to Water on Graphene

“…20 Other studies focused on characterizing the hydrophobic effect on electrically doped graphene layers. 21 MD simulations with the TIP4P water model 22 were used in Ref. 23 to examine the desalination performance of graphene, while water and ion transport through graphene pores was investigated in Refs.…”

Mass-flow-rate-controlled fluid flow in nanochannels by particle insertion and deletion

“…(1a) and (1b) y is the direction perpendicular to the wall, ρ B and µ are the bulk density and bulk viscosity of the fluid, L y is the channel height with walls located at ±L y /2, and L z is the channel depth. Several methods have been developed to induce planar Poiseuille flow in MD [24][25][26][27][28][29][30][31][32][33][34][35][36][37][38][39]. The most commonly used technique is the body force method (BFM) [24].…”

“…Typically a constant force is applied (BFM-C). Recently, a BFM based on constraint dynamics has been developed [25]. Termed Gaussian dynamics by the authors and here as BFM-GD, the method maintains a desired momentum and thus controls the mass flow rate in the system.…”