"Glass" has a unique and distinct meaning in physics—one that refers not just to the transparent material we associate with window glass. Instead, it refers to any system that looks solid but is not in true equilibrium and continues to change extremely slowly over time. Examples include window glass, plastics, metallic glasses, spin glasses (i.e., magnetic systems), and even some biological and computational systems.

When a liquid is cooled very quickly—a process called quenching—it doesn't have time to organize into a crystal but becomes stuck in a disordered state far from equilibrium. Its properties—like stiffness and structure—slowly evolve through a process called "aging."

Now, a research team from the Institute of Theoretical Physics of the Chinese Academy of Sciences has proposed a new theoretical framework for understanding the universal aging behavior of glassy materials.

The study reveals a fundamental mechanism that governs how glasses—from simple spin systems to complex network glasses such as amorphous silica—slowly evolve over time.

To understand the aging process, the researchers developed a generalized trap model (GTM) grounded in the material's energy landscape: a multidimensional map of all possible configurations and the energy barriers that separate them. According to the GTM, aging is driven by activated hopping across these energy barriers. A universal distribution of barrier heights, incorporating crucial finite-size corrections, governs the system's slow, nonequilibrium dynamics.

The theory predicts that during nonequilibrium aging, the system undergoes "weak ergodicity breaking" at a temperature higher than the conventional glass transition temperature. In statistical physics, "ergodic" refers to a system that explores all possible configurations consistent with its energy. In contrast, the term "ergodicity breaking" refers to an equilibrium system becoming trapped in a subset of possible states, unable to explore all configurations. Weak ergodicity breaking occurs in nonequilibrium systems and describes a system that continues to evolve but remains correlated with its initial configuration even after prolonged aging.

By applying the GTM to four distinct models, including the random energy model (a spin glass), the Weeks-Chandler-Andersen model (a simple atomic glass), and amorphous silica (a network glass), the researchers demonstrated that glass aging behavior follows universal mathematical laws. A key finding is that the logarithmic decay of the two-time correlation function, a hallmark of aging, is directly linked to the finite size of "activation clusters," or groups of particles that rearrange together during the aging process.

In the Weeks-Chandler-Andersen model, this insight allowed the researchers to extract a static length scale from the nonequilibrium dynamics, extending its observable growth range from a mere factor of two to three to a full order of magnitude. This provides strong supporting evidence for the random first-order transition (RFOT) theory, a leading theory of the glass transition.

This work provides a unified phase diagram that describes both ergodic and weakly non-ergodic phases in spin and structural glasses, offering a powerful tool for understanding these ubiquitous yet complex materials. These findings have implications not only for materials science but also for other complex systems, such as protein dynamics and even the training of deep learning algorithms, where similar slow relaxation processes are observed.