Utrecht physicists create equilibrium glassy phase from rod-shaped particles

[Lead]
Glass appears to be a solid, but in theory it sometimes behaves more like an extremely slow liquid. Physicists in Utrecht now show that glass-like structures can also exist in equilibrium, which is something many theories say should be impossible.

[Intro]
The bottom parts of medieval window panes, such as those in old cathedrals, are often thicker than the top. Has the material slowly flowed downward over the centuries, and does this mean that glass actually flows?

This is a persistent myth, and the explanation lies in the way glass was produced in the Middle Ages. Because window panes were made by hand, their structure was often irregular and contained thinner and thicker parts. The panes were usually installed in the frame with the thicker side at the bottom, which made them more stable.

Still, the story touches on a real physics question. What glass actually is, a solid or a very slow liquid, turns out to be more difficult to answer than it seems.

Equilibrium
Researchers at Utrecht University have now created a glass-like state that is in thermodynamic equilibrium. According to common theories, such a state should not really be able to exist. “A glass and an equilibrium state exclude each other in many people’s minds,” says Thijs Besseling, first author of the study.

A chaotic state
Even within physics there is still a lot that is unclear about the glass phase. Glass has no neat internal ordering, like a crystal. At the same time, the molecules also do not move freely past each other, as they would in a liquid.

In textbooks, glass is therefore often described as a material that is just out of equilibrium: a liquid that has been cooled so quickly that the atoms get stuck in a chaotic state. And in nature, such a state should not be stable, at least according to theory.

Rod glass
In their experiment, the researchers did not study ordinary glass made of atoms. They created a model system of colloids. Colloids are larger than atoms, but they show similar physical behaviour. Because of their size, colloids can easily be followed under a microscope, which is not possible for atoms.

Researchers often use spherical colloids, but this time the team chose rod-shaped particles. Rods can not only move sideways, but they can also rotate. The team created a state in which the positions of the rods are disordered and ‘frozen’, as in glass, while the particles themselves can still rotate. This state turns out to be energetically favourable, which means the system naturally returns to it.

A push with an electric field
With an external electric field, the researchers temporarily push the rod glass toward a crystal structure. According to theory, that should be a more stable final state. But as soon as the field is switched off again, the material returns to the glass-like state. This shows that the system is not simply stuck in some intermediate state, but that it is actually in equilibrium.

Testing with simulations

“The claim that a glass-like state can be in equilibrium is quite controversial,” Besseling explains. That is why the researchers collaborated with scientists specialised in simulations, led by Laura Filion. They used computer simulations of a simplified model and observed the same behaviour there: at higher densities a disordered state is more stable than the crystalline state. “So, we see it both in a complex experiment and in a simple computer model. That gives us confidence in the results.”

What does this mean for ‘real’ glass?

The study does not prove that window glass is suddenly in equilibrium. Atomic systems behave differently from colloidal model systems. However, the research does show that the shape of particles can play an important role. Many molecules are not perfectly spherical: some have an elongated shape and can rotate. The results suggest that these rotations may be important in the glass transition, which is an element that is missing in many existing theories.

A fundamental understanding of the glass transition would make it possible to accurately describe and predict the properties and lifetime of disordered materials, thereby making materials development much faster and more efficient. This could have an impact on the development of, among others, batteries, coatings, biomaterials, and industrial manufacturing processes. It could also be relevant to biology and medicine, because biological processes such as cell growth also exhibit glass-like dynamics.

Thijs Besseling is Principal Scientist at imec the Netherlands and part-time researcher at the Soft Condensed Matter and Biophysics group at the Debye Institute of Nanomaterials Science. For this research he collaborated with Berend van der Meer, Bing Liu, Laura Filion, Arnout Imhof and Alfons van Blaaderen.