Glassy Dislocation Dynamics in 2D Colloidal Dimer Crystals

Although glassy relaxation is typically associated with disorder, here we report on a new type of glassy dynamics relating to dislocations within 2D crystals of colloidal dimers. Previous studies have demonstrated that dislocation motion in dimer crystals is restricted by certain particle orientations. Here, we drag an optically trapped particle through such dimer crystals, creating dislocations. We find a two-stage relaxation response where initially dislocations glide until encountering particles that cage their motion. Subsequent relaxation occurs logarithmically slowly through a second process where dislocations hop between caged configurations. Finally, in simulations of sheared dimer crystals, the dislocation mean squared displacement displays a caging plateau typical of glassy dynamics. Together, these results reveal a novel glassy system within a colloidal crystal.

Figure 1

Micrographs illustrating a local perturbation experiment. Fivefold- and sevenfold-coordinated defects created by the perturbation are marked with pink and blue dots, respectively. (a) A single spherical intruder particle in a small grain is dragged by 1 LC by using optical tweezers. (b) Two dislocations marked by arrows glide to the grain edge and are absorbed by the grain boundary (solid closed curve). (c) The intruder particle remains stationary after the tweezers are turned off. The vacancy left behind has no topological charge and is stable. (d)–(f) In a large grain (only the central part is shown here) the dislocations marked by arrows are obstructed by the particles in red. When the tweezers are turned off, the blocked dislocations return along their glide paths, recombining and causing the intruder particle to recoil.

Figure 2

Recoil distance $R$ as a function of glide distance $Z$ to the nearest grain boundary. The error bars represent uncertainty in particle position. The distribution of unrestricted glide distances $\rho (Z)$ (dashed gray curve) is reprinted from Ref. 8.

Figure 3

Average number of defects versus time for large ($>700$ particles) degenerate crystals and crystals of spheres. Error bars represent the standard error of the mean. The solid line is the best fit exponential for the decay in the sphere data ($R^2=0.99$); the dashed line is the best fit sum of an exponential and a logarithmic decay for the dimer data ($R^2=0.99$).

Figure 4

Defect trajectories in simulated crystals under shear (large arrows). (a) Dislocations in a crystal of spheres glide along straight paths in response to the applied shear. (b) Dislocations in a degenerate crystal follow crooked trajectories as they hop between caged configurations. (c) MSD of individual dislocations in crystals of spheres and degenerate crystals. Error bars are smaller than the markers. The solid curve is the best fit for $\mathrm{MSD}=⟨\Delta r^2⟩=u^2{t}^2+2Dt$.