Bereiche | Tage | Auswahl | Suche | Aktualisierungen | Downloads | Hilfe
CPP: Fachverband Chemische Physik und Polymerphysik
CPP 23: Modeling and Simulation of Soft Matter III
CPP 23.4: Vortrag
Mittwoch, 20. März 2024, 10:15–10:30, H 0107
Efficient construction of high-dimensional neural network potentials for the Strecker synthesis — •Alea Miako Tokita1,2, Timothée Devergne3, A Marco Saitta3, and Jörg Behler1,2 — 1Theoretische Chemie II, Ruhr-Universität Bochum, Germany — 2Research Center Chemical Sciences and Sustainability, Research Alliance Ruhr, Germany — 3Institut de Minéralogie, de Physique des Matériaux et de Cosmochimie, Sorbonne Université, Paris, France
High-Dimensional Neural Network Potentials (HDNNPs) provide potential energy surfaces with the accuracy of electronic structure calculations at strongly reduced computational costs. This enables extended molecular dynamics simulations of large systems such as organic molecules in solution. The construction of a HDNNP for such systems is not a trivial task since they have a vast configuration space which needs to be efficiently sampled. An example for such a complex system is the first step of the classic Strecker-cyanohydrin mechanism for glycine synthesis in water from formaldehyde and hydrogen cyanide. Here, we present a systematic construction of a HDNNP for this system as a showcase for molecular chemistry in solution.
Keywords: Machine Learning; Atomistic Simulations; Strecker Mechanism