Application of Vibration-Transit Theory of Liquid Dynamics to the Brillouin Peak Dispersion Curve

TL;DR

This paper applies vibration-transit (V-T) theory to model the Brillouin peak dispersion in liquids, showing that vibrational contributions alone accurately match experimental and simulation data for liquid sodium.

Contribution

It introduces a V-T theory-based model for the Brillouin peak dispersion that does not rely on decay process assumptions, differing from mode coupling theory.

Findings

Vibrational contribution alone matches experimental data

Model accurately predicts Brillouin peak dispersion in liquid Na

No adjustable parameters needed in the vibrational model

Abstract

The Brillouin peak appears in the dynamic structure factor S(q,w), and the dispersion curve is the Brillouin peak frequency as function of q. The theoretical function underlying S(q,w) is the density autocorrelation function F(q,t). A broadly successful description of time correlation functions is provided by mode coupling theory, which expresses F(q,t) in terms of processes through which the density fluctuations decay. In contrast, vibration-transit (V-T) theory is a Hamiltonian formulation of monatomic liquid dynamics in which the motion consists of vibrations within a many-particle random valley, interspersed with nearly instantaneous transits between such valleys. Here, V-T theory is applied to S(q,w). The theoretical vibrational contribution to S(q,w) is the sum of independent scattering cross sections from the normal vibrational modes, and contains no explicit reference to decay processes. For a theoretical model of liquid Na, we show that the vibrational contribution with no adjustable parameters gives an excellent account of the Brillouin peak dispersion curve, as compared to MD calculations and to experimental data.