# SKM 2023 – wissenschaftliches Programm

## Bereiche | Tage | Auswahl | Suche | Aktualisierungen | Downloads | Hilfe

# O: Fachverband Oberflächenphysik

## O 78: Heterogeneous Catalysis and Surface Dynamics II

### O 78.2: Vortrag

### Donnerstag, 30. März 2023, 10:45–11:00, TRE Phy

Adaptive Experimental Designs for the Unsupervised Exploration of Reaction Kinetic Phase Diagrams — •Frederic Felsen, Karsten Reuter, and Christoph Scheurer — Fritz-Haber-Institut der MPG, Berlin, Germany

In heterogeneous catalysis the interplay of various elementary processes taking place at the catalytic interface gives rise to complex kinetic phase diagrams [1]. Characteristic for such phase diagrams are regimes of smooth kinetic behavior separated by phase transitions going along with drastic changes in effective kinetics. Generally it is difficult to reconstruct the kinetic regime topology of a given process without a detailed understanding of the underlying reaction mechanism. Here, we tackle this issue by approximating the reaction kinetics using statistically robust effective rate laws [2]. Locally fingerprinting the apparent behavior of a chemical reaction, we can generate point-wise regime labels as a basis for estimates of the position of the phase transitions. In order to achieve this in a data efficient way, we propose an iterative adaptive design algorithm, bridging between the classical theory of experimental design [3] and modern statistical learning and optimization. Using a microkinetic model, we illustrate the potential of this approach by investigating the kinetic phase diagram of CO oxidation over RuO_{2}.

[1] V. Zhdanov et al., Surf. Sci. Rep. 20, 113, (1994).

[2] F. Felsen et al., Chem. Eng. J. 433, 134121, (2022).

[3] V. Fedorov, Theory of Optimal Experiments (1972).