Dresden 2014 – wissenschaftliches Programm

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

O: Fachverband Oberflächenphysik

O 10: Solid-liquid Interfaces I

O 10.6: Vortrag

Montag, 31. März 2014, 11:45–12:00, WIL A317

Development of a Neural Network Potential for Protonated Water Clusters — •Suresh Kondati Natarajan, Tobias Morawietz, and Jörg Behler — Lehrstuhl für Theoretische Chemie, Ruhr Universität Bochum, D-44780 Bochum, Germany

Protonated water clusters are important model systems for the dissociation of water, proton transport and acids in aqueous solution. Many classical force fields are not applicable to the dissociation of water or proton transport due to their inability to describe the making and breaking of bonds. Recently, high-dimensional neural networks (NNs) [1-2] have been introduced as an efficient way to represent interatomic potentials based on electronic structure calculations. Here, we present a NN potential for protonated water clusters with up to eight water molecules. The results are discussed and compared to density functional theory data.

[1] J. Behler and M. Parrinello, Phys. Rev. Lett., vol. 98, p.146401, 2007.

[2] J. Behler, Phys. Chem. Chem. Phys., vol. 13, p.17930, 2011.