 |
Center for Nonlinear Studies |
Electrochemical interfaces play a critical role in a wide spectrum of applications ranging from energy storage in batteries to separation processes and industrial chemical reactions. Despite their importance, routine simulations of interfacial phenomena at solid-liquid interfaces are challenging due to the complex interplay between concentration-dependent solid-ion, solid-solvent, ion-solvent, and solvent-solvent interactions that dictate the macroscopic properties of the electrochemical devices. This is further complicated by the inability of conventional molecular dynamics (MD) simulations to faithfully represent a realistic electrochemical system due to accumulation of solvated species at the interface and their resulting depletion from the bulk. Simple systems like NaCl, consisting of only monovalent ionic species, have been studied extensively both experimentally and computationally, primarily due to their relative simplicity. However, more complex electrolytes such as those containing divalent ions, for example Na2SO4, have not been extensively studied due in part to the difficulty of obtaining experimental spectroscopic data and the long time and length scales required to obtain converged DFT-quality properties.
In this presentation, I will describe my work related to developing a simple iterative method, termed the iterative constant chemical potential MD approach (iCµMD), for achieving constant chemical potential conditions that integrates machine learning interatomic potentials (MLIPs) which provide DFT-quality simulations of interfacial phenomena at arbitrary concentrations. I show that the iCµMD approach can quickly converge to a target bulk concentration for graphene/graphite-electrolyte interfaces using classical force fields and MLIPs. I will also discuss my recent work in utilizing MLIPs to understand the effect of Na2SO4 and the formation of the electric double layer in graphene-electrolyte interfaces, which are critical for energy storage applications and electrocatalysis. Here, I will discuss how the non-linear sum-frequency generation spectroscopy can be used to understand the structure of the electrical double layer and how the concentration and organization of ions at the interface affects the ordering of interfacial water molecules. This work shows the broad applicability of MLIPs for modeling aqueous or other solvated systems and their impact on interfaces.
Click to join via Zoom
Meeting ID: 821 3903 8740
Passcode: 359699---