Flow quantization and nonequilibrium nucleation of soft crystals

Abstract

We consider dense systems consisting of ultrasoft, overlapping particles under shear and transport flow, by employing a multiscale simulational approach that combines multiparticle-collision dynamics for the solvent particles with standard molecular dynamics for the solute. We find that the nucleation rates of supercooled liquids can be dramatically accelerated via the shear-induced formation of an intermediate string pattern, which disaggregates after the cessation of shear, leading to the emergence of three-dimensional fcc order. Furthermore we expose these cluster crystals to Poiseuille flow and we establish the emergence of a quantized flow pattern, in which both the height and the width of the fluid stream display well-defined plateaus as a function of the applied pressure gradient. The resulting velocity profiles of the solvent closely resemble plug flow. We explain the emergence of the plateaus by successive fluidization of crystalline layers adjacent to the channel walls and discuss the dependence of the discrete flow on the cluster aggregation parameter. Cluster crystals thus emerge as novel systems with applications on nano- and microfluidic devices, allowing the manipulation of flow in a precisely controlled way.