Self-diffusion process in water: Spatial picture of single-particle density fluctuations

Abstract

A computer simulation methodology with which to study the single-particle dynamics in complex molecular liquids is presented. Molecular dynamics simulations of liquid water are performed in the temperature range of 238–473 K using the polarizable point charge (PPC) potential. The self part of the van Hove density–density correlation function is calculated. Using the Gaussian approximation of the van Hove function the mean self-diffusion coefficient for the PPC potential is calculated. The singularity temperature for supercooled PPC water, Ts=218 K, estimated from the self-diffusion data appears to agree well with most estimates for real water. In order to elucidate the spatial picture of the single-particle molecular density in this complex liquid and its time evolution, we explicitly resolve the self van Hove function in the local frame of water molecules. The self-diffusion tensor is introduced and numerically evaluated from this spatial (separation and direction dependent) self van Hove function. The fluctuations of the single-particle molecular density in liquid water appear to be spatially anisotropic (nonspherical). At low temperatures these dynamical heterogeneities in liquid water tend to increase.