Maximally Random Jammed Packings
- Maximally Random Jammed Packings are strictly jammed, disordered configurations defined by a variational order metric and minimal isostatic contacts.
- They display hyperuniformity with suppressed infinite-wavelength density fluctuations and universal quasi-long-range correlations characterized by precise power laws.
- MRJ packings extend to diverse particle shapes, mixtures, and confined geometries, offering insights into transport properties, electromagnetic behavior, and computational challenges.
Maximally random jammed (MRJ) packings represent the most disordered configurations of nonoverlapping particles that are strictly jammed, i.e., mechanically stable against all collective and global deformations. MRJ packings form a paradigmatic model for disordered, rigid, athermal materials—glassy states poised at the margins of mechanical and dynamic stability. Central features include isostaticity, hyperuniform suppression of large-scale density fluctuations, universal quasi-long-range correlations, and dimension-dependent exponents quantifying their structural singularities. The MRJ paradigm generalizes beyond monodisperse spheres to arbitrary convex shapes, polydisperse mixtures, and high-dimensional Euclidean and confined geometries.
1. Variational Definition, Mechanical Stability, and Isostaticity
The MRJ state is defined as the strictly jammed configuration that minimizes a specified scalar order metric ψ, quantifying disorder. Strict jamming implies incompatibility with any collective displacement or volume-nonincreasing boundary deformation; no strictly jammed packing can be continuously unjammed without overlaps. Maxwell counting for frictionless spheres in dimension yields an isostatic contact number per particle, , such that the total number of mechanical contacts satisfies in the thermodynamic limit. For monodisperse spheres in , the MRJ state occurs at and a packing fraction ; almost every backbone sphere participates in exactly the minimal number of contacts required for mechanical stability (Jiao et al., 2011, Torquato et al., 2010).
MRJ states are uniquely distinguished among jammed packings by being simultaneously mechanically rigid and maximally disordered with respect to ψ. The variational principle subject to jamming, and (if desired) fixed , singles out the MRJ configuration from the generically nonunique set of jammed packings at each density (Jiao et al., 2011).
2. Universality of Hyperuniformity and Quasi-Long-Range Correlations
A defining feature of MRJ packings is disordered hyperuniformity: the anomalous suppression of infinite-wavelength density (number or local-volume-fraction) fluctuations. Hyperuniformity is verified by the structure factor vanishing as 0, typically with a power law 1 where 2; the corresponding real-space pair correlations decay as 3 for large 4. For monodisperse spheres, 5 in 6, leading to 7, and in general, 8 for 9 (Maher et al., 2023). These quasi-long-range (QLR) correlations are universal across particle shapes (spheres, ellipsoids, superdisks, superballs, truncated tetrahedra), sizes (monodisperse and binary), and even in confined geometries (Zachary et al., 2010, Zachary et al., 2011, Jiao et al., 2010, Maher et al., 2021, Chen et al., 2014, Chen et al., 2015).
Hyperuniformity is consistently observed in the two-phase medium (local-volume-fraction fluctuations), even when the point pattern of centers is not hyperuniform due to polydispersity or anisotropy (Zachary et al., 2011). The suppression of large-scale fluctuations is tied to the rigidity of the void space—loops of contacting particles enclosing all voids in finite, highly regular shapes (Zachary et al., 2011).
3. Marginal Stability, Quasicontacts, and Weak-Force Distributions
MRJ packings are poised at a double marginal: statically at the isostatic contact number, and dynamically at the onset of unstable collective rearrangements. This marginality is manifest in two signature power-laws:
- Quasicontacts: The distribution of near contacts—pairs at gap 0—exhibits a singularity in the pair correlation function, 1 for 2, with 3 in low dimensions. The count of quasicontacts within gap 4 scales as 5.
- Weak contact forces: The distribution of normal contact forces follows 6 for 7, with 8 in 9.
Dynamic marginality relates these exponents by the Wyart stability bound: 0; in MRJ packings, this is nearly saturated (Kallus et al., 2014).
Quasicontacts, negligible for static stability, become dynamically relevant by providing the shortest paths for contact closure upon infinitesimal perturbations. In high dimensions, the abundance parameter 1 of quasicontacts grows faster than that of contacts, altering prefactors in theoretical density estimates and dominating in certain Bravais lattice cases (Kallus et al., 2014).
4. Structural Characteristics and Statistical Descriptors
MRJ packings display a suite of structural features:
- Isostatic backbone and rattlers: The jammed network consists of a mechanically rigid backbone interspersed by rattlers (locally unjammed particles), with the rattler fraction decreasing with dimension and anisotropy. Ideal MRJ states in the infinite system limit may be rattler-free (Maher et al., 2023, Torquato et al., 2010).
- Order metrics: The MRJ state minimizes translational and/or orientational order metrics—translational order parameter 2, bond-orientational order 3, or general ψ (Torquato et al., 2010, Jiao et al., 2011).
- Voronoi anticorrelations: Two-point correlation functions of Minkowski functionals (volume, surface area, mean curvature) of Voronoi cells exhibit strong anticorrelations, consistent with the suppression of large-scale density fluctuations. Large Voronoi cells are statistically paired with small ones, sharply distinguishing MRJ from unjammed liquids or uncorrelated Poisson processes (Klatt et al., 2015).
- Packing spectrum: MRJ packings anchor the lower boundary of the φ–order parameter plane among all jammed packings, with the continuous spectrum extending upwards to fully ordered crystals; the full set of jammed packings at varying density displays protocol dependence, except for the uniquely defined MRJ point (Jiao et al., 2011, Torquato et al., 2010).
5. Generalization to Shape, Mixtures, and Geometries
The MRJ paradigm—and associated structural features—extends robustly:
- Nonspherical shapes: MRJ packing fractions and contact statistics depend drastically on shape deformation parameters (e.g., superdisk/superball exponent 4). For superballs, 5 increases monotonically and nonanalytically with 6, and jammed networks become hypostatic, supported by nongeneric local arrangements of contacts (Jiao et al., 2010, Maher et al., 2021).
- Binary and polydisperse mixtures: Binary hard-disk MRJ states in 2D display a hyperuniformity exponent 7 maximized (up to 8) for certain size ratios 9 (Maher et al., 2024). In 3D, polydispersity preserves hyperuniformity in the medium even if number density fluctuations are not hyperuniform (Zachary et al., 2011).
- High dimensions and Bravais lattices: In 0, quasicontacts significantly increase the theoretical prefactor for packing density, and in certain isostatic Bravais lattices, dominate the asymptotic scaling exponent (Kallus et al., 2014).
- Confinement: Under plane or wall confinement, MRJ packings retain global hyperuniformity while exhibiting layering, heightened rattler fractions, and discontinuous transitions in response to geometric frustration (Chen et al., 2015).
6. Transport, Electromagnetic, and Physical Property Implications
The disordered hyperuniformity of MRJ packings engenders exceptional transport, diffusive, and electromagnetic properties compared to conventional amorphous or crystalline media:
- Transport: Rigorous upper/lower bounds and universal scaling laws for permeability and mean survival time in porous MRJ media show dramatically reduced values, reflecting the uniformity and smallness of the voids (Klatt et al., 2017, Klatt et al., 2016). The principal relaxation time and trapping constants approach those of periodic lattices as the local void geometry is regularized by jamming.
- Wave propagation: MRJ packings are nearly dissipationless for long-wavelength electromagnetic waves due to hyperuniformity—Im1 in the quasistatic limit. Stealthy hyperuniform variants could even yield finite-wavelength complete photonic gaps (Klatt et al., 2017).
- Topology of the void space: MRJ packings possess a narrow, compact-support pore-size distribution set by local loops in the contact network, with void regions sharply bounded by the jamming and saturation constraints (Zachary et al., 2011).
7. Algorithms, Protocols, and Open Challenges
Protocols for generating MRJ packings include event-driven compression (Lubachevsky–Stillinger), sequential (or adaptive) linear programming (Torquato–Jiao), and Monte Carlo adaptive-shrinking-cell methods for non-spherical particles. The protocol-dependence of jammed states outside the MRJ point has been demonstrated; only the MRJ state specified by variational minimization over order metrics is universally defined (Jiao et al., 2011, Torquato et al., 2010, Chen et al., 2014).
Critical slowing down complicates generation of large, truly strictly jammed, rattler-free MRJ packings. Ensuring exact jamming and hyperuniformity for 2 remains a fundamental challenge, intimately connected to the saturation of collective rearrangements near the jamming threshold (Atkinson et al., 2016, Maher et al., 2023).
Further, capturing the full spectral density—including the amplitude of small-3 linear scaling and the direct prediction of void-space statistics from first principles—remains an open theoretical frontier (Zachary et al., 2011).
References (arXiv ID)
- Marginal stability, quasicontacts, and power-law force/gap distributions: (Kallus et al., 2014)
- Non-universality and variational definition of MRJ: (Jiao et al., 2011)
- Hyperuniformity and universal QLR correlations: (Maher et al., 2023, Zachary et al., 2010, Zachary et al., 2011, Jiao et al., 2010, Maher et al., 2021, Chen et al., 2014, Chen et al., 2015)
- Protocol-dependence and the φ–order map: (Jiao et al., 2011, Torquato et al., 2010)
- Voronoi statistics and Minkowski correlations: (Klatt et al., 2015)
- Transport and EM property implications: (Klatt et al., 2016, Klatt et al., 2017)
- Rattlers, algorithmic obstacles, and scaling: (Atkinson et al., 2016, Maher et al., 2024)
- Binary and polydisperse MRJ packings: (Maher et al., 2024, Zachary et al., 2011)
- Geometric-structure theory for nonspherical shapes: (Tian et al., 2016)
- Analytic RCP/MRJ density via isostaticity: (Zaccone, 2022)