Karl Franzens University Graz
Graz University of Technology
|
|
| |
Karl Franzens University Graz | Graz University of Technology | |
Out of the Crystalline Comfort Zone: ML-Empowered Modelling of Operando Energy Conversion Systems Operando spectroscopies and microscopies reveal a highly dynamic behavior of interfaces in energy conversion systems. Insufficient insight and the concomitant inability to control or exploit the corresponding strong structural and compositional modifications centrally limits the development of performance catalysts, electrolyzers or batteries required for a sustainable energy supply for our society. Predictive-quality modeling and simulation has become a major contributor to accelerated design all across the materials sciences, not least through powerful computational screening approaches. Current first-principles based methodology is nevertheless essentially unable to address the substantial, complex and continuous morphological transitions at working interfaces. I will review this context from the perspective of first-principles based multiscale modeling, highlighting that the fusion with modern machine learning approaches is key to tackle the true complexity of working systems. Approaches pursued by our group thereby aim at maximum data efficiency by exploiting physical models wherever possible or through active learning that only queries data on demand. One cornerstone of such agile data-empowered multiscale modeling is the iterative training of first-principles surrogate models in the form of machine learned interatomic potentials. At an orders of magnitude lower computational cost, these potentials now enable first simulations beyond idealized crystalline interfaces and allowing enhanced sampling methods to increasingly replace traditional human input based on chemical intuition. |