Estimating random close packing in polydisperse and bidisperse hard spheres via an equilibrium model of crowding

Carmine Anzivino, Mathias Casiulis, Tom Zhang, Amgad Moussa, Stefano Martiniani, Alessio ZacconeThe Journal of Chemical Physics 2022 · DOI · arXiv

Abstract

We show that an analogy between crowding in fluid and jammed phases of hard spheres captures the density dependence of the kissing number for a family of numerically generated jammed states. We extend this analogy to jams of mixtures of hard spheres in d = 3 dimensions and, thus, obtain an estimate of the random close packing volume fraction, $\phi_{RCP}$, as a function of size polydispersity. We first consider mixtures of particle sizes with discrete distributions. For binary systems, we show agreement between our predictions and simulations using both our own results and results reported in previous studies, as well as agreement with recent experiments from the literature. We then apply our approach to systems with continuous polydispersity using three different particle size distributions, namely, the log-normal, Gamma, and truncated power-law distributions. In all cases, we observe agreement between our theoretical findings and numerical results up to rather large polydispersities for all particle size distributions when using as reference our own simulations and results from the literature. In particular, we find $\phi_{RCP}$ to increase monotonically with the relative standard deviation, $s_\sigma$, of the distribution and to saturate at a value that always remains below 1. A perturbative expansion yields a closed-form expression for $\phi_{RCP}$ that quantitatively captures a distribution-independent regime for $s_\sigma < 0.5$. Beyond that regime, we show that the gradual loss in agreement is tied to the growth of the skewness of size distributions.

Introduction

In statistical mechanics, hard spheres serve as the fundamental model for condensed matter. While it is well-established that monodisperse spheres pack randomly to a volume fraction of approximately 64% ($\phi \approx 0.64$), real-world materials—such as sand, colloids, or industrial powders—are typically "polydisperse," consisting of particles with varying sizes.

A simulated jammed state of a polydisperse hard-sphere mixture.

Predicting the maximum density, or Random Close Packing (RCP), of these mixtures is a significant challenge. Jamming is inherently a non-equilibrium process, making it difficult to describe using standard equilibrium thermodynamics. Existing models are often empirical or rely on computationally expensive simulations. This paper introduces a unified, analytical framework to predict how $\phi_{RCP}$ evolves as a function of size variation.

Methodology: The Equilibrium Analogy

The researchers propose a novel "equilibrium analogy," hypothesizing that the structural crowding in a jammed state mirrors the crowding in a highly dense equilibrium fluid.

  1. The Isostaticity Criterion: In three dimensions, a stable packing of frictionless spheres requires a minimum average of 6 contacts per particle (the "kissing number," $z=6$). The authors define RCP as the densest possible state that maintains this isostatic condition.
  2. Liquid-State Theory: By utilizing established Equations of State (EOS) for liquids—such as Percus-Yevick and Carnahan-Starling—the model relates the volume fraction ($\phi$) to the coordination number $z$.
  3. Generalization to Mixtures: These equations are extended to accommodate both discrete binary mixtures and continuous size distributions (e.g., log-normal, Gamma).
Phase diagram showing the relationship between kissing number (z) and volume fraction (phi), identifying the RCP point and the MRJ line.

The study validated these theoretical predictions against numerical simulations using Lubachevsky-Stillinger and Torquato-Jiao algorithms, as well as historical benchmarks from literature.

Key Insights

  • Bridging Fluid and Jammed States: The mapping of local crowding physics from the liquid branch onto non-equilibrium jammed states allows the application of liquid-state physics to solve complex jamming problems.
  • Small-Polydispersity Universality: For relative standard deviations ($s_\sigma$) below 0.5, the specific shape of the size distribution is secondary. The packing density follows a universal parabolic increase dependent primarily on the variance.
  • Saturation and Skewness: As polydispersity increases, $\phi_{RCP}$ increases monotonically but plateaus around 0.97–0.98, never reaching 1.0 due to persistent interstitial voids. Discrepancies at very high polydispersity are attributed to the "skewness" (asymmetry) of the size distribution, which standard EOS models do not fully capture.

Results and Implications

The model demonstrates excellent quantitative agreement with simulations for binary mixtures across a wide range of diameter ratios (1.2 to 5). Furthermore, the authors provide a simplified rational function that allows for rapid estimation of density shifts ($\Delta \phi_{RCP}$) in mixtures with moderate polydispersity.

The theoretical consistency of the model is highlighted by its ability to identify RCP as a unique point in the ensemble of jammed states, marking the end of the isostatic plateau before the onset of crystallization toward a face-centered cubic (fcc) arrangement.

Limitations and Future Directions

While robust, the model's accuracy diminishes when size distributions are extremely skewed. Future work may involve developing more sophisticated equations of state that incorporate higher-order moments of the distribution. Additionally, while this study focuses on spheres, the "crowding model" could potentially be adapted for non-spherical particles, such as ellipsoids or irregular grains, which are common in practical applications.

Summary updated

Abstract

We show that an analogy between crowding in fluid and jammed phases of hard spheres captures the density dependence of the kissing number for a family of numerically generated jammed states. We extend this analogy to jams of mixtures of hard spheres in d = 3 dimensions and, thus, obtain an estimate of the random close packing volume fraction, ϕ<sub>RCP</sub>, as a function of size polydispersity. We first consider mixtures of particle sizes with discrete distributions. For binary systems, we show agreement between our predictions and simulations using both our own results and results reported in previous studies, as well as agreement with recent experiments from the literature. We then apply our approach to systems with continuous polydispersity using three different particle size distributions, namely, the log-normal, Gamma, and truncated power-law distributions. In all cases, we observe agreement between our theoretical findings and numerical results up to rather large polydispersities for all particle size distributions when using as reference our own simulations and results from the literature. In particular, we find ϕ<sub>RCP</sub> to increase monotonically with the relative standard deviation, s<sub>σ</sub>, of the distribution and to saturate at a value that always remains below 1. A perturbative expansion yields a closed-form expression for ϕ<sub>RCP</sub> that quantitatively captures a distribution-independent regime for s<sub>σ</sub> < 0.5. Beyond that regime, we show that the gradual loss in agreement is tied to the growth of the skewness of size distributions.
BibTeX
@article{anzivino2022estimating,
  author = {Carmine Anzivino and Mathias Casiulis and Tom Zhang and Amgad Moussa and Stefano Martiniani and Alessio Zaccone},
  title = {{Estimating random close packing in polydisperse and bidisperse hard spheres via an equilibrium model of crowding}},
  journal = {The Journal of Chemical Physics},
  year = {2022},
  doi = {10.1063/5.0137111},
}
Figure