Randium: A minimal model of universal viscous liquid dynamics

The paper introduces "Randium," a minimal two-dimensional lattice model that demonstrates how universal features of viscous liquid dynamics, such as time-temperature superposition and parabolic scaling, emerge from simple nearest-neighbor rearrangements without requiring elasticity-induced facilitation.

The Gist

Imagine you have a giant, crowded dance floor filled with thousands of dancers. When the music is fast and loud (high temperature), everyone moves freely, bumping into each other, spinning, and changing partners easily. The crowd flows like a liquid.

But as the music slows down and the room gets colder, the dancers start to get stuck. They get trapped in little circles with their neighbors, shuffling in place but unable to leave. Eventually, the whole floor freezes into a solid block, even though the dancers are still jiggling slightly. This is what happens to liquids when they turn into glass.

Scientists have long noticed something strange: no matter what the liquid is made of (water, oil, or complex chemicals), once it gets cold enough to become glassy, they all behave in almost the exact same way. They slow down, get stuck, and relax in a very specific pattern.

This paper introduces a new, super-simple computer model called Randium to explain why this happens.

The "Randium" Game

Think of Randium as a giant checkerboard (a grid).

• The Pieces: Instead of black and white checkers, every square has a "particle" with a random personality type.
• The Rules: The only thing that matters is how much a particle likes its four immediate neighbors. Some pairs get along great (low energy), while others hate each other (high energy). These "likes" and "dislikes" are assigned randomly, like drawing numbers from a hat.
• The Action: The only way the system changes is if two neighbors swap places. They only swap if the new arrangement makes them happier (or if they are brave enough to try a swap that makes them slightly unhappy, hoping to get lucky later).

There are no complex physics rules here. No long-range forces, no elasticity, and no complicated chemistry. Just a grid, random neighbors, and a temperature setting.

What Happens in the Game?

When the "temperature" in the game is high, particles swap places constantly. The system relaxes quickly, just like a warm liquid.

But as the temperature drops, something magical and universal happens:

1. Getting Stuck: Particles try to swap, but they often realize the new neighbors are worse than the old ones. So, they swap back. They are "trapped" in their little cages.
2. The Chain Reaction: Occasionally, a swap happens that does work. This small change might make a nearby particle's neighbors suddenly seem more friendly. That neighbor can now move, which helps its neighbor move.
3. The Cascade: This creates a chain reaction. A small group of particles starts moving together, breaking out of their cages. This is called dynamic facilitation.

Why Is This Important?

The paper shows that this simple game of "random swaps on a grid" perfectly mimics the behavior of real, complex liquids turning into glass.

• The Shape of Time: When scientists measure how long it takes for real liquids to relax, the curve looks like a specific mathematical shape (a "stretched exponential"). Randium produces the exact same shape without being programmed to do so.
• The "Universal" Curve: The authors compared their game results to real-world data from dozens of different chemicals (from water to oils). Randium's results fit right on top of the real data.
• No "Elasticity" Needed: Some scientists thought that long-range "elastic" forces (like a rubber band pulling from far away) were necessary to explain why glass forms. Randium proves they are wrong. You don't need long-range forces; you just need local neighbors helping each other out.

The Big Picture

The paper argues that the complex, messy physics of real glass-forming liquids can be boiled down to this simple idea: Local cooperation.

Just like a crowd of people where one person moving creates space for the next person to move, the "glassy" behavior of liquids emerges naturally from simple, local rules. Randium is a "minimal model"—it strips away all the unnecessary details to show that the core engine of glass formation is surprisingly simple.

In short: You don't need a complex recipe to make glass behave like glass. You just need a grid of neighbors who occasionally help each other escape their traps. That simple rule is enough to explain the universal behavior of liquids turning into solids.

Technical Summary

Technical Summary: Randium – A Minimal Model of Universal Viscous Liquid Dynamics

Problem Statement
When liquids are cooled without crystallizing, their dynamics slow dramatically, eventually solidifying into an amorphous glass. Experimental evidence indicates that chemically distinct glass-forming liquids share universal features in their structural relaxation spectra and temperature dependence. While various theoretical frameworks exist (e.g., kinetic constraint models, trap models, elastic models), a central question remains: Can a simple, energetically coarse-grained model, devoid of long-range elastic facilitation, reproduce these generic dynamics? Specifically, is long-range elasticity required to explain universal viscous liquid behavior, or can it emerge from local interactions alone?

Methodology
The authors introduce Randium, a minimal model of viscous liquids implemented on a two-dimensional square lattice with periodic boundary conditions.

• System Definition: The system consists of $N = L^2$ particles, each assigned a "type" from a set of $M$ types. The energy of a microstate is defined by the sum of nearest-neighbor interactions.
• Energetics: The interaction matrix $I$ between particle types is symmetric, with elements drawn from a standard normal (Gaussian) distribution. This replaces the complex energy landscape of real molecular systems with a spatially organized network of random energies, consistent with observations from atomistic molecular dynamics simulations.
• Dynamics: The model evolves via Monte Carlo (MC) simulations using nearest-neighbor swap attempts and Boltzmann acceptance criteria. A particle swap represents a local fundamental collective motion (a rearrangement of a small group of particles). The dynamics are intrinsic, independent of system size, and governed by a single control parameter: temperature ($T$).
• Simulation Scale: Results are presented for a system size of $L=192$ ($N=36,864$), utilizing parallelized algorithms on GPUs to access the long time-scales characteristic of viscous liquids.

Key Contributions

1. Model Construction: The paper constructs a minimal model that satisfies four criteria: Gaussian thermodynamics, spatial locality of flows, intrinsic dynamics, and a single control parameter.
2. Exclusion of Elasticity: Unlike competing theories emphasizing long-range elasticity-induced facilitation, Randium explicitly excludes this mechanism. It demonstrates that universal dynamics can emerge solely from local rearrangements and energetic traps.
3. Emergent Phenomena: The model shows that dynamic facilitation, dynamical heterogeneity, and free-energy barriers emerge naturally from the rules, rather than being imposed via kinetic constraints (as in KMC) or pre-defined trap distributions (as in trap models).

Results

• Relaxation Spectra: At high temperatures, the overlap function $Q(t)$ decays nearly exponentially. At low temperatures, the decay resembles a stretched exponential with an exponent of $1/2$, consistent with dielectric experiments and one-dimensional kinetic constraint models.
• Time-Temperature Superposition: The relaxation curves at different low temperatures collapse onto a single universal curve when scaled by a characteristic relaxation time $\tau$. This universal curve is distinct from a simple stretched exponential and matches experimental data from depolarized dynamic light scattering on chemically distinct molecular liquids.
• Temperature Dependence: The structural relaxation time $\tau$ follows a parabolic scaling law, $\tau \propto \exp(J^2(\beta - \beta_0)^2)$, in agreement with Elmatad, Chandler, and Garrahan's findings for molecular systems. The model estimates an inverse glass-transition temperature $\beta_g = 2.16$ and an Angell fragility index $m=43$, typical of molecular glass-formers.
• Dynamical Heterogeneity: As temperature decreases, the size of cooperatively rearranging regions increases. The characteristic length scale $\xi$ grows, and the relaxation time scales exponentially with this length scale ($\tau \propto \exp(\xi)$). This leads to a decoupling of timescales (Stokes-Einstein breakdown) at low temperatures.
• Comparison to Other Models: Randium's relaxation behavior differs from trap models (which exhibit fat-tailed dynamics due to a lack of facilitation) and aligns more closely with the Random Barrier Model (RBM) in terms of mean squared displacement scaling.

Significance and Claims
The paper claims that Randium provides an intrinsic, energetically grounded route to understanding universal viscous liquid dynamics. By reproducing the generic relaxation spectra, time-temperature superposition, and parabolic scaling of real systems without invoking long-range elastic facilitation, the model suggests that elasticity is not a necessary condition for these universal features.

The authors position Randium as a "stepping-stone framework" that bridges realistic molecular models and highly idealized approaches (such as the Random Barrier Model or kinetically constrained models). They conjecture that this class of models, governed by similar physics, includes variations with different lattice connectivities (e.g., 3D cubic lattices) and alternative interaction distributions, implying that the precise realization of the system has little influence on the universal dynamical behavior aside from trivial scaling factors. The work offers a path toward connecting complex molecular dynamics to more fundamental, analytically tractable models.