Scientists led by a physicist at the University of Oregon have taken a major step in solving an enduring mystery that we encounter every time we look through a window or stare at a phone screen.
For centuries, scientists have studied how and why glass forms from a molten state into a rigid, solid material, while still maintaining an amorphous or disordered internal structure like a liquid.
UO physicist Eric Corwin and a team of current and former students have now created on a computer the first “ideal glass,” a material where the molecules are packed together as tightly and stably as possible while still being amorphous.
Scientists have long theorized that such a material should exist but have been unable to create it physically or mathematically.
Understanding how to make this ideal glass could open the door to new kinds of materials with unique properties, such as glass that could withstand high heat and pressure. Since material properties determine how products can be made, that, in turn, could lead to new, more efficient manufacturing processes for everything from golf clubs to engines.
Corwin and his colleagues reported their results in the journal Physical Review Letters .
To physicists, glassy materials encompass more than what most people typically think of as glass: windows, bottles, cellphone screens and eyeglasses. Anything solid made of amorphous arrangements of molecules — like many plastics, metallic glasses and even some biological materials — qualifies as a glass, scientifically speaking.
In his lab, Corwin focuses on how shape affects function, essentially how the shape of a material’s molecules affects its physical characteristics. For example, he said, consider a cup of water. When cooled from its liquid state, water forms ice, an orderly crystalline solid at the molecular level. In contrast, glass molecules form a rigid structure but are actually disordered, like grains of sand on a beach.
“If you look at glass at a molecular level, you would see that the molecules are arranged amorphously,” Corwin said. “They’re kind of random. They’re all pushed up against one another, but there’s no structure.”
In a crystalline solid, molecules occur in a regular, predictable pattern, but in an amorphous solid, randomly dispersed molecules are stuck in place rather than moving as they would in a liquid.
“So this is the big question about glasses,” he said. “How do you get stability, mechanical stability, in a system that is totally amorphous, that looks like a liquid?”
In their study, Corwin and his colleagues took up a challenge laid down in 1948 by Walter Kauzmann, a Princeton University chemist. He theorized that as glass cools to extremely cold temperatures, it would eventually arrive at an ideal state in which the molecules are packed as tightly as possible. As the molecules reach this state, the material would behave much like a crystalline solid and have enhanced properties such as a higher melting point, flexibility, or exceptional resistance to breaking under stress.
Such ideal glass does not exist in nature, so there is nothing for scientists to study directly. However, Corwin’s team decided to create one through mathematical modeling on the UO high-performance computer cluster.
“We thought maybe we can just jump to it,” Corwin said. “We can construct the best possible structure.” They started by creating an arrangement of molecules shaped like round disks.
Corwin’s team drew insight from the structure of a two-dimensional crystal, in which each disk is surrounded by six neighboring disks and is in contact with all its neighbors, like cells in a honeycomb. They then developed a method to preserve the structure of disks perfectly packed against one another, while entirely removing the repeated crystalline structure.
Their work demonstrated that these new structures are the densest possible configurations of a given set of disks. They confirmed this ideal glass by comparing the model’s physical properties to the known qualities of crystalline solids. By looking at how it responded to pressure, bending, melting, and other features, they showed that their model acted more like a crystalline than an amorphous solid.
“The conclusion is that our structure mechanically behaves identically to a crystal, even though it is completely amorphous,” Corwin said.
Corwin and his team will continue to expand their work into three-dimensional space. Eventually, their results could improve manufacturing processes and potentially allow for better understanding of materials like metallic glasses, Corwin suggested. Metallic materials with a disordered structure, versus the orderly structure of a conventional metal or alloy, have a variety of interesting and useful properties. They can be very strong and resistant to being deformed, and they can be melted and used for injection molding.
Currently, such materials are difficult to make because they must be cooled rapidly from a liquid to a solid state, limiting their usefulness.
“If we could develop a much better understanding of the glass transition and understand what makes an alloy (a combination of metals) better or worse at forming a metallic glass, we could design alloys that you could cool much more slowly. And then you could really do things. You could mold a car engine, you could mold a jet fighter. It would be revolutionary,” Corwin said.
— By Nick Houtman
This research was supported by the Simons Foundation. The research team also included UO doctoral student Viola Bolton-Lum, as well as R. Cameron Dennis and Peter Morse, who previously received their doctoral degrees in Corwin’s lab.
Physical Review Letters
Ideal Glass and Ideal Disk Packing in Two Dimensions
2-Feb-2026