Induced Charge Anisotropy: A Hidden Variable Affecting Ion Transport through Membranes

Abstract

Conventional molecular simulations are generally incapable of capturing the kinetics of solute transport through ultra-selective membranes. Here, we use a recently developed path-sampling technique called jumpy forward-flux sampling to accurately and efficiently compute long passage times for sodium and chloride ions traversing a nanoporous graphitic membrane. We also demonstrate that an ion's passage is not only impeded by its partial dehydration but also by a negative restraining force due to the charge anisotropy that it induces at its rear.

Watching Water, Sodium, and Chloride Passing through a Graphitic Pore

“…3 Strong deviations from macroscopic predictions in carbon nanotubes occur because of the atomistic smoothness of the walls of these structures. 4 The hydrogen-terminated pore simulated by Malmir et al 2 does not share this smoothness. The estimated salt rejection of the simulated membrane is >99.99%, meaning that >10,000 water molecules pass through for every ion.…”

“…In this issue of Matter, Malmir et al have taken a step in this direction by introducing methods that model permeation through an ultra-thin membrane in impressive detail. 2 Analyzing permeation of saltwater through a membrane using atomistic molecular simulations seems superficially simple, but the very limited timescales imposed by molecular dynamics (MD)-on the order of nanoseconds to microseconds-place severe constraints on the number of molecular events that can be observed. This restricts the use of these methods in simulating stochastic dynamic events, such as ion permeation.…”

“…The challenge is particularly acute for a membrane with a high level of salt rejection, because ions in the solution pass through the membrane orders of magnitude slower than water molecules. Malmir et al 2 tackled this timescale challenge in two ways. First, they modeled a membrane of just three layers of graphite made porous by including a single sub-nm pore through the layers.…”

“…Crudely, the flux through a membrane of thickness L scales as 1/L, so the choice of an ultra-thin membrane maximizes this flux. By applying hydrostatic pressure in their MD simulations, Malmir et al 2 were able to directly observe water permeation from a 1.5 M NaCl solution.…”

“…6,7 When this is the case, it can be challenging to extrapolate insight from ultra-thin membranes to more experimentally realistic structures. Simulations are best suited to membranes that are high ordered, 6,7 as in the single pore simulated by Malmir et al, 2 but more realistic polymeric membranes are disordered and amorphous. Efforts have been made to simulate membranes of this kind, but these face even greater challenges associated with timescales and sampling.…”