Kinetics of Domain Growth and Aging in a Two-Dimensional Off-lattice System

TL;DR

This study uses molecular dynamics simulations to analyze phase separation, domain growth, and aging in a 2D Lennard-Jones system, revealing different growth regimes and aging behaviors depending on temperature and phase coexistence.

Contribution

It provides detailed insights into the kinetics of domain growth and aging in a 2D off-lattice model, including crossover phenomena and scaling laws, which were not previously characterized in such detail.

Findings

Domain growth follows t^{1/2} in vapor-liquid phase separation.

A crossover to slower growth occurs at low temperatures, indicating solid-like states.

Aging autocorrelation exhibits power-law decay with a well-estimated exponent.

Abstract

We have used molecular dynamics simulations for a comprehensive study of phase separation in a two-dimensional single component off-lattice model where particles interact through the Lennard-Jones potential. Via state-of-the-art methods we have analyzed simulation data on structure, growth and aging for nonequilibrium evolutions in the model. These data were obtained following quenches of well-equilibrated homogeneous configurations, with density close to the critical value, to various temperatures inside the miscibility gap, having vapor-"liquid" as well as vapor-"solid" coexistence. For the vapor-liquid phase separation we observe that $\ell$, the average domain length, grows with time ($t$) as $t^{1/2}$, a behavior that has connection with hydrodynamics. At low enough temperature, a sharp crossover of this time dependence to a much slower, temperature dependent, growth is identified within the time scale of our simulations, implying "solid"-like final state of the high density phase. This crossover is, interestingly, accompanied by strong differences in domain morphology and other structural aspects between the two situations. For aging, we have presented results for the order-parameter autocorrelation function. This quantity exhibits data-collapse with respect to $\ell/\ell_w$, $\ell$ and $\ell_w$ being the average domain lengths at times $t$ and $t_w$ ($\leq t$), respectively, the latter being the age of a system. Corresponding scaling function follows a power-law decay: $~\sim (\ell/\ell_w)^{-\lambda}$, for $t\gg t_w$. The decay exponent $\lambda$, for the vapor-liquid case, is accurately estimated via the application of an advanced finite-size scaling method. The obtained value is observed to satisfy a bound.