Quantifying structural dynamic heterogeneity in a dense two-dimensional equilibrium liquid.

We investigate local structural fluctuations in a model equilibrium fluid with the aim of better understanding the structural basis of locally heterogeneous dynamics identified in recent simulations and experimental studies of glass-forming liquids and other strongly interacting particle systems, such as lipid membranes, dusty plasmas, interfacial dynamics of crystals, the internal dynamics of proteins, etc. In particular, we utilize molecular dynamics simulation methods to study a single component Lennard-Jones condensed material at constant temperature in two dimensions over a range of densities covering both liquid and crystalline phase regimes. We identify three distinct structural classes of particles by examining the immediate neighborhood of individual particles relying on a solid-angle based tessellation technique. The area distribution of the neighborhoods reveals cages having hexagonal, pentagonal, and square symmetries. Pentagonal cells appear to be the predominant motif in the liquid phase, while the solid phase is dominated by hexagonal cells, as in the case of a perfect crystal. An examination of the spatial organization of particles belonging to each structural class further indicates that finite-size clusters of the hexagonal and pentagonal particle populations arise within both liquids and solids, and the size of these clusters grows in a complementary way as a function of density. Both particle populations form percolation clusters in the liquid-crystal coexistence regime. Interestingly, the populations of particles with different local structures, defined by the arrangement of neighboring particles, are found to maintain different diffusivities, as computed from the velocity autocorrelation function for each type of particle for all densities studied. Our analysis provides a new conceptual framework for understanding the structural origin of dynamical heterogeneity in soft materials.