Unifying Atoms and Colloids near the Glass Transition through Bond-Order Topology

Am excited to tell you about a recent Physical Review Letter involving the Soft Materials Group at ETHZ and myself, in which we compared a simulated model metallic binary glass structure to an experimental binary colloidal system.

In atomistic simulation, one can produce a glassy structure by performing molecular dynamics to first obtain a high temperature equilibrium liquid. The temperature is then lowered at some fixed rate fast enough to avoid the equilibrium phase transition to a crystalline solid. This is easy in simulation because only very short time-scales can be simulated. As the temperature reduces, the meta-equilibrium of the under-cooled liquid is entered and at some sufficiently low temperature, Tg, the atoms "freeze" or arrest over the characteristic time-scale of the simulation, producing an amorphous solid. The temperature at which this occurs depends on the temperature quench rate, where a slower quench rate entails a lower Tg.

The above phenomenon occurs in all sorts of materials over a broad range of time and length scales, ranging from atoms (metallic, ionic and covalent systems), to molecules, to colloids and granular systems. However for these latter (larger scale) systems, where thermal fluctuations can be negligible, density replaces the role of temperature. In these systems, instead of lowering the temperature, one increases the density, and a loosely packed flowing system eventually transits to a type of "jammed" glassy state. An everyday example is sand on the beach - watery sand lowers the density of the grains, and the combined mixture flows, however when the water leaves the sand, the grain density increases and flow ceases.

In our paper, we investigate through a particle-scale structural comparison, the relationship between density (in an experimental mechanically compressed athermal binary colloidal system) and temperature (in a simulated thermally quenched model thermal binary alloy). We derive a mapping between the two systems and find that the latter arrests at a temperature above the simulated thermal glass transition temperature, Tg, thus having a structure more akin to the under-cooled liquid than the amorphous solid. I find this result interesting, since it demonstrates, that (in this case) the athermal colloidal system cannot probe the more relaxed structural motifs that the simulated thermal system has access to through activated structural excitations.

For more details see: Unifying Atoms and Colloids near the Glass Transition through Bond-Order Topology, L. Stricker, P. M. Derlet, A. F. Demir?rs, H. R. Vutukuri, and J. Vermant, Phys. Rev. Lett. 132, 218202 (2024).