https://doi.org/10.1140/epjst/e2016-60120-1
Regular Article
Comparison of iterative inverse coarse-graining methods
1 Eduard-Zintl-Institut für Anorganische und Physikalische Chemie and Center of Smart Interfaces, Technische Universität Darmstadt, Alarich-Weiss-Strasse 10, 64287 Darmstadt, Germany
2 Institut für Mathematik, Johannes Gutenberg Universität Mainz, 55099 Mainz, Germany
a e-mail: vandervegt@csi.tu-darmstadt.de
Received: 15 April 2016
Revised: 25 June 2016
Published online: 10 October 2016
Deriving potentials for coarse-grained Molecular Dynamics (MD) simulations is frequently done by solving an inverse problem. Methods like Iterative Boltzmann Inversion (IBI) or Inverse Monte Carlo (IMC) have been widely used to solve this problem. The solution obtained by application of these methods guarantees a match in the radial distribution function (RDF) between the underlying fine-grained system and the derived coarse-grained system. However, these methods often fail in reproducing thermodynamic properties. To overcome this deficiency, additional thermodynamic constraints such as pressure or Kirkwood-Buff integrals (KBI) may be added to these methods. In this communication we test the ability of these methods to converge to a known solution of the inverse problem. With this goal in mind we have studied a binary mixture of two simple Lennard-Jones (LJ) fluids, in which no actual coarse-graining is performed. We further discuss whether full convergence is actually needed to achieve thermodynamic representability.
© EDP Sciences, Springer-Verlag, 2016
