(605a) Bridging the Gap between Electrokinetic Experiments and Molecular Dynamics Simulations

The Electric Double Layer (EDL) that forms at the interface between a charged solid and an aqueous solution is of paramount importance in many scientific fields ranging from geology and material science to biology and electrochemistry. A detailed understanding of the EDL and its dependence on the solid material, solution composition, and pH is thus invaluable for improvements in for example catalysis, batteries or colloid sciences. However, no single experimental technique can measure the entire structure and dynamics of the EDL. Instead, various techniques can provide partial pieces to the puzzle, but only seldomly can these techniques and their results be combined unambiguously. Some experimental results even contradict each other. Molecular Dynamics (MD) simulations could provide unambiguous insight into the EDL. They are however, only as accurate in predicting EDL properties as permitted by their atomic interaction models and the experimental data used to fit these. Unfortunately, the experimental data itself is often ambiguous, especially for interfaces. For example, different electrokinetic techniques at the same conditions frequently yield different ζ-potential values. The difficulty is thus obtaining unambiguous experimental data and using this data to tune the atomic interaction models. In this work, we present such data and an approach to tune atomic interaction models, while at the same time minimizing modeling errors. Specifically, we use the fact that solution composition and pH for which the ζ-potential is zero, are uniquely defined no matter the electrokinetic technique, e.g. electroosmosis or streaming current. In fact, following the Helmholtz-Smoluchowski theory, the relationship between the measured quantity, flow velocity or electric current, and ζ-potential is linear. Hence, also the measured quantities are zero when the ζ-potential equals zero, leaving the relationship between the measured quantities and ζ-potential independent of other fluid properties such as viscosity. This renders the solution composition and pH at which the ζ-potential equals zero as ideal to tune atomic interaction models. We then show that a simple scaling of the ion-surface Lennard-Jones cross interactions can already yield very good agreement between experiments and simulations in predicting the conditions at which the ζ-potential is zero. Essentially, by tuning these interactions we alter the adsorption behavior of ions. Particularly, we can control the number of specifically adsorbed ions and by increasing or reducing these, the experimental conditions for a zero ζ-potential can be matched. We strongly believe that this simple strategy in tuning atomic interaction models, in combination with carefully designed experiments, can significantly expand our toolbox to analyze the EDL, and both, experiments and simulations, can benefit each other in a symbiotic relationship.