All mineral surfaces in contact with aqueous solution become charged and attract electrolyte ions from the solution phase to form a characteristic interfacial charge distribution, known as the electrical double layer (EDL). When two charged mineral/electrolyte interfaces are brought into proximity to each other, the overlap of their electrical double layers perturbs electrostatic potential and leads to the release or uptake of charges; a phenomenon called the charge regulation. As a result, the interfacial charge and potential distributions vary with the separation between overlapping double layers.<br/>            The stability of particles in the suspension depends on the magnitude of electrostatic interactions between particles - double-layer forces. Double-layer forces govern rates of many dynamic phenomena, such as aggregation, coagulation, coalescence, flocculation, adhesion, sorption, and deposition. Our ability to predict and model double-layer forces relay on the accuracy of the mathematical models of the charged interfaces and forces arising due to their overlap. Our limited ability to probe these interfaces hamper the development of realistic theoretical models.<br/>Here, we show how molecular dynamics, ab-initio, and Monte Carlo simulations can be used to parameterize meso- and macroscopic models of the overlapping EDLs. First, we used ab-initio molecular simulations to determine intrinsic proton and ion-affinities for surface sites exposed by mineral surfaces in contact with an aqueous solution. Second, we used molecular dynamics simulations to determine the spatial and temporal distribution of ions near the surface and interfacial fluid properties, such as dielectric constant and viscosity. Finally, we show how machine learning surrogates based on the deep neural networks and ensemble methods can replace the computationally expensive solvers of an emerging set of equations. These surrogates significantly reduce the computational cost and allow an ultrafast analysis of a broad range of experimental data.

Molecularly informed charge regulation models of overlapping electrical double layers and their machine learning surrogates

ACS Spring 2022

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.