English

Hierarchical Lattice Models of Hydrogen Bond Networks in Water

Statistical Mechanics 2018-06-19 v2 Materials Science Soft Condensed Matter Chemical Physics

Abstract

We develop a graph-based model of the hydrogen bond network in water, with a view towards quantitatively modeling the molecular-level correlational structure of the network. The networks are formed are studied by the constructing the model on two infinite-dimensional lattices. Our models are built \emph{bottom up}, based on microscopic information coming from atomistic simulations, and we show that the predictions of the model are consistent with known results from ab-initio simulations of liquid water. We show that simple entropic models can predict the correlations and clustering of local-coordination defects around tetrahedral waters observed in the atomistic simulations. We also find that orientational correlations between bonds are longer ranged than density correlations, and determine the directional correlations within closed loops and show that the patterns of water wires within these structures are also consistent with previous atomistic simulations. Our models show the existence of density and compressibility anomalies, as seen in the real liquid, and the phase diagram of these models is consistent with the singularity-free scenario previously proposed by Sastry and co-workers (Sastry et al, PRE 53, 6144 (1996)).

Keywords

Cite

@article{arxiv.1709.08703,
  title  = {Hierarchical Lattice Models of Hydrogen Bond Networks in Water},
  author = {Rahul Dandekar and Ali A. Hassanali},
  journal= {arXiv preprint arXiv:1709.08703},
  year   = {2018}
}

Comments

17 pages, published version