Collective self-sorting on a chip

Abstract

We harness two established ingredients for collective demixing: differential speed and curvature to create a self-sorting device. In binary mixtures, motility differences drive spontaneous spatial segregation, while confinement geometry determines how rapidly and strongly this demixing develops. Using particle based simulations, we systematically identify the geometrical conditions that promote efficient segregation and use these results to guide the design of a finite sorting architecture. We then translate these physical mechanisms into a sequence of curved microfluidic units that progressively amplify the separation of the two species and direct them toward distinct collection regions. Experiments with binary Quincke-roller mixtures confirm that an initially mixed suspension progressively demixes as it propagates through the device, leading to strong enrichment downstream. Our results demonstrate how collective active demixing can be converted into a functional continuous sorting strategy, providing a route toward autonomous microfluidic separation based on particle motility and confinement geometry.