(248j) Overturning Intrinsic Temporal Limits in Atomic Scale Simulation of Reactive Systems

Myriad research topics in chemical engineering and materials science belong to the category of “reactive systems.” Fundamental understanding of these systems is achieved most definitively and most efficiently with atomic level modeling. State-of-the-art methods based in Molecular Dynamics (MD) such as ab initio MD (AIMD), reactive MD, and hybrid MD/Kinetic Monte Carlo (HMD/KMC) provide such spatial resolution. However, simulation past nanosecond timescales is not possible due to the significant computational demand. Here, a new HMD/KMC algorithm is introduced that is capable of atomistic simulation of reactive systems to arbitrarily long timescales. This is achieved by placing the burden of timekeeping on the KMC portion of the methodology, as opposed to the current strategy of using MD for the official timeclock. Despite being heuristic, this algorithm has been used to accurately reproduce dynamic and equilibrium behavior of multiple test systems. Homogeneous gas- and liquid-phase systems provide facile calculation of reference data through the application of kinetic theory. Accordingly, the HMD/KMC method shows quantitative agreement of molecular concentrations across timescales previously achievable only through continuum modeling (e.g. tens of seconds). This algorithm can be used to explain phenomena of reactive systems for which experimentation and continuum modeling cannot provide definitive answers. For example, device failure in lithium-ion batteries is acutely sensitive to degradation of the Solid Electrolyte Interphase (SEI). Due to its highly heterogeneous structure and slow degradation reactions, neither continuum models nor current atomistic simulations can readily describe these processes. Simultaneously, these phenomena occur at the nanoscale over significant periods of time, making experimental measurement difficult and expensive, at best. With this algorithm, an accurate description at the appropriate spatial resolution will provide crucial insights into how and why the SEI degrades. This talk will discuss the structure and capabilities of this HMD/KMC algorithm. Through presentation of the validation systems, it will be shown that quantitatively accurate atomistic simulations of reactive systems are no longer restricted to short time scales.