Berlin 2024 – scientific programme
Parts | Days | Selection | Search | Updates | Downloads | Help
MM: Fachverband Metall- und Materialphysik
MM 44: Developement of Calculation Methods II
MM 44.6: Talk
Wednesday, March 20, 2024, 17:15–17:30, C 264
Kernel Charge Equilibration: Machine Learned Interatomic Potentials With Full Long-Range Electrostatics — •Martin Vondrak1, Johannes T. Margraf1,2, and Karsten Reuter1 — 1Fritz-Haber-Institut der MPG, Berlin — 2University of Bayreuth
Machine learning (ML) interatomic potentials have recently been shown to bridge the gap between accurate first-principles methods and computationally cheap empirical potentials. This is achieved by learning a mapping between a system’s structure and its physical properties. To this end, state-of-the-art models typically represent chemical structures in terms of local atomic environments. This inevitably leads to the neglect of long-range interactions (most prominently electrostatics) and non-local phenomena (e.g. charge transfer), resulting in significant errors in the description of polar molecules and materials (particularly in non-isotropic environments). To overcome these issues, we recently proposed an ML framework for predicting charge distributions in molecules termed Kernel Charge Equilibration (kQEq) [1]. Here, atomic charges are derived from a physical model using environment-dependent atomic electronegativities. In this contribution, strategies for creating kQEq interatomic potentials are discussed, including the combination of short-ranged Gaussian Approximation Potentials with kQEq.
[1] M. Vondrak, K. Reuter, and J.T. Margraf, J. Chem. Phys. 159, 054109 (2023).
Keywords: Machine learning; Charge Equilibration; Partial Charges; Interatomic Potentials; Dipole Moments