By Ashley Yeager

An artist shapes a piece of molten glass. Credit: Sarah Weiser / The Herald

The nature of a glass is a debate that runs far deeper than whether one appears half empty or half full.

For years scientists have known that the atoms and molecules in glasses arrange themselves randomly, like the non-uniform pattern of particles in a liquid.

What’s puzzling, though, is that while the molecules in a glass are arranged like in a liquid, the overall structure is solid.

“We’re close to understanding why certain materials do not crystallize, but it’s still an open challenge to explain how these materials then become viscous and form glasses,” says Patrick Charbonneau, a theoretical chemist at Duke.

In other words, scientists do not yet have an accurate theory to describe both the liquid and solid characteristics of glass.

Instead, there’s just an old joke that says there are more explanations about how glasses form and behave than there are theorists to propose the ideas.

molecular structure of glass

Click the graphic to learn more about the molecular intricacies of glass. Credit: Jonathan Corum, The New York Times.

Using new computer simulations, Charbonneau has begun to wipe away some of the less influential theories, and with his collaborators Atsushi Ikeda, Giorgio Parisi and Francesco Zamponi, he has identified serious flaws in one of the leading explanations describing the nature of glass.

The results of the research appear in the Oct. 28 issue of the journal Physical Review Letters.

In the study, Charbonneau tested the glass theories using simulations designed to look at a material’s atomic and molecular interactions in up to 12 dimensions. If scientists can make sense of how glass particles interact in these higher dimensions, then “we’ll be in a better position for understanding two and three-dimensional glass formation,” he says.

One of the leading theories of glass works, sometimes, but for it to hold more water, it needs to be modified so that its predictions match what the 12-dimensional simulations are showing.

If scientists can’t corroborate the predictions and simulated results, they may need to completely overhaul all they think they know about the nature of glasses — empty or full.

Citation: Charbonneau et al. Phys. Rev. Lett. 107, 185702 (2011).
DOI: 10.1103/PhysRevLett.107.185702