If you freeze a liquid fast enough, it becomes a glass, something that is structurally similar to liquid but incapable of flow. This concept holds true even for metals. These vitrified metals, or metallic glasses, are at the frontier of materials science research, but much about them remains poorly understood.

From a practical standpoint, metallic glasses combine the advantages, and avoid many of the problems, of both ordinary metals and glasses, two classes of materials that have very wide ranges of applications.

From a structural perspective, metallic glasses lack regular atomic packing associated with conventional crystalline metals. This means they also lack the defects found in conventional crystalline metals and instead offer many unique properties combining extremely high strength, high hardness, good wear resistance, high corrosion resistance.

However, even after decades of effort since the birth of metallic glass in 1960s, scientists’ understanding about the atomic structure of metallic glass is still limited to very local scale. Unlike crystals, which are organized in repeating patterns that extend in every direction, glasses can demonstrate clusters of order among the nearest, neighboring atoms. For this reason, no two glasses, produced under the same conditions, are identical at the atomic level. How these local clusters of order extend to efficiently fill up the three dimensional space without crystalline symmetry has remained unanswered.

Using a variety of advanced techniques including x-ray diffraction, tomography, and molecular dynamics simulations, a team including the Geophysical Laboratory's Qiaoshi “Charles” Zeng, was able to create a model that describes both the observed short-range order in metallic glasses and encompasses long-range structural details. The research team’s findings suggest that the structure of metallic glass is created when the material’s atoms percolate, or filter down through holes in the structure, while in the liquid phase, and the percolating cluster becomes rigid at the glass transition temperature. The model has considerable implications for understanding glass properties and the origin of the liquid-to-glass transition.

Figure caption: (A). Bulk metallic glasses with different shapes and sizes prepared by cupper mold casting. (B) The overall atomic structure of a Cu46Zr54 system simulated by molecular dynamics. (C) The percolating network of Cu-centered icosahedra clusters hidden inside the sample.

Scientific Area: