Parts | Days | Selection | Search | Updates | Downloads | Help

MO: Fachverband Molekülphysik

MO 10: Cluster

MO 10.6: Talk

Tuesday, March 15, 2011, 11:45–12:00, MER 02

High-dimensional neural network potential-energy surface for water clusters — •Vikas Sharma, Tobias Morawietz, and Jörg Behler — Lehrstuhl für Theoretische Chemie, Ruhr-Universität Bochum, D-44780 Bochum, Germany.

Studying the structural and dynamical properties of water clusters provides valuable insights into many important processes, e.g. in biology, materials science, and atmospheric chemistry. However, a prerequisite for the accurate description of water clusters is the availability of a reliable interatomic potential. In the present work artificial Neural Networks (NNs) trained to first-principles calculations are used to represent the multi-dimensional potential-energy surface of water clusters of different sizes. The quality of the NN potential is investigated by studying various stationary points as well as the vibrational frequencies of several benchmark clusters. The obtained results are compared to electronic structure data.