Monte Carlo simulation of kinetically slowed down phase separation

Abstract

Abstract. Supercooled colloidal or molecular systems at low densities are known to form liquid, crystalline or glassy drops, which may remain isolated for a long time before they aggregate. This paper analyses the properties of this large time window, and how it can be tackled by computer simulation. We use singleparticle and virtual move Monte Carlo simulations of short-range attractive spheres which are undercooled to the temperature region, where the spinodal intersects the attractive glass line. We study two different systems and we report the following kinetic behavior. A low-density system is shown to exhibit universal linear growth regimes under single-particle Monte Carlo correlating the growth rate to the local structure. These regimes are suppressed under collective motion, where droplets aggregate into a single large disordered domain. It is shown that the aggregation can be avoided and linear regimes recovered, if long-range repulsion is added to the short-range attraction. The results provide an insight into the behavior of the virtual move algorithm generating cluster moves according to the local forcefields. We show that different choices of maximum Monte Carlo displacement affect the dynamical trajectories but lead to the same kinetically slowed down or arrested states.

Self-assembly of ellipsoidal particles at fluid-fluid interfaces with an empirical pair potential

“…For this reason we begin with many independent random initial configurations and average over their time evolutions, to imitate experimental procedures where the particles are initially spread onto an interface and then allowed to self-assemble 230 [50]. This procedure is known in the literature as a 'rapid' or 'instantaneous' quench [53,54,55].…”

“…Given the severe unreliability of the quadrupole approximation for small sep- 55 arations and the prohibitive computational cost of solving the Young-Laplace equation numerically for many particles in order to simulate the structure formation process, we prefer a different approach. Our approach consists of developing an empirical pair potential for ellipsoidal particles, fitting it to exact numeri-4 cal calculations of ellipsoid pairs at different separation distances and relative 60 orientations, which goes beyond the rather inaccurate but typically assumed quadrupolar pair potential [33] by taking into account the ellipsoidal geometry.…”