Berlin 2024 – scientific programme
Parts | Days | Selection | Search | Updates | Downloads | Help
MM: Fachverband Metall- und Materialphysik
MM 62: Developement of Calculation Methods III
MM 62.1: Talk
Thursday, March 21, 2024, 15:45–16:00, C 264
Phase transitions in radial distribution biased Molecular Dynamics simulations — •Lars Dammann1,2, Patrick Huber1,2, and Robert H. Meißner1,3 — 1Hamburg University of Technology, Hamburg, Germany — 2Deutsches Elektronen-Synchrotron DESY, Hamburg, Germany — 3Helmholtz-Zentrum Hereon, Geesthacht, Germany
Molecular Dynamics (MD) simulations are crucial for the study of phenomena at the atomic level. However, simulating the transition from liquid to solid states remains a challenge due to the energy barrier between phases, that makes the observation of spontaneous phase transitions in unbiased simulations computationally infeasible. Radial distribution functions (RDFs) provide valuable insights into atomic structures but are not unambiguous and therefore hard to interpret. I will present an algorithm that biases molecular dynamics simulations to reproduce a target RDF while minimizing the amount of information with which the original force field is biased. This is achieved through the application of the principle of maximum relative entropy. Following this principle, it is possible to use the structural information about the system contained in the RDF while conserving as much information as possible about the atomic interactions contained in the original force field. I will demonstrate that this strategy can be used to promote liquid-solid phase transitions in MD simulations. In the future, this method could potentially aid in the investigation of complex phase transitions, the analysis of metastable states and for the interpretation of structural information from experimental data.
Keywords: Biased Molecular Dynamics simulations; Radial distribution functions; Phase transitions; Maximum relative entropy; Structural modelling