Speaker
Description
Modeling complex energy materials such as solid-state electrolytes (SSEs) realistically at the atomistic level strains the capabilities of state-of-the-art theoretical approaches. On one hand, the system sizes and simulation time scales required are prohibitive for first-principles methods like density-functional theory (DFT). On the other hand, parameterizations for empirical potentials are often not available and these potentials may ultimately lack the desired predictive accuracy.
Fortunately, modern machine learning (ML) potentials are increasingly able to bridge this gap, promising first-principles accuracy at a much-reduced computational cost. However, the local nature of these ML potentials typically means that long-range contributions arising, e.g., from electrostatic interactions are neglected. Such interactions can be large in polar materials like electrolytes, however, though it is not necessarily clear how this affects the prediction of specific observables (such as ion mobilities).
To shed some light on this issue, we investigated the effect that the locality assumption of ML potentials has on lithium mobility and defect formation energies in the solid electrolyte Li7P3S11 [1]. We found that neglecting long-range electrostatics is unproblematic for the description of lithium transport in the isotropic bulk. In contrast, (field-dependent) defect formation energies are only adequately captured by a hybrid potential combining ML and a physical model of electrostatic interactions.
In order to obtain such electrostatic models capable of describing polarization and charge transfer, we subsequently developed the Kernel Charge Equilibration (kQEq) approach, which expands the classical QEq model with a machine-learned, environment dependent electronegativity [2]. This enables an accurate description of charge distributions in molecules and materials, at low computational cost. Currently, this is being further developed to allow for the simultaneous fitting of long- and short-ranged components of the potential energy.
[1] C.G. Staacke, H.H. Heenen, C. Scheurer, G. Csányi, K. Reuter, and J.T. Margraf, ACS Appl. Energy Mater. 4, 12562 (2021).
[2] C.G. Staacke, S Wengert, C. Kunkel, G. Csányi, K. Reuter, and J.T. Margraf,
Mach. Learn. Sci. Tech. 3, 015032 (2022).
| Abstract Number (department-wise) | TH 03 |
|---|---|
| Department | TH (Reuter) |
