Skip to content Skip to navigation

Hierarchical organization of chiral colloidal

Liquid–liquid
phase separation in bulk proceeds through the continuous coalescence
of droplets until the system undergoes complete phase separation. But when
colloids, nanoparticles or proteins are confined to interfaces, surfaces or
membranes, their interactions differ fundamentally from those mediated by
isotropic solvents, and this results in significantly more complex phase behaviour. We have shown that liquid–liquid
phase separation in monolayer membranes composed
of two dissimilar chiral colloidal rods gives rise to thermodynamically stable
rafts that constantly exchange monomeric rods with the background reservoir to
maintain a self-limited size. Our
observations demonstrate a robust membrane-based pathway for the assembly of monodisperse membrane clusters that is
complementary to existing methods for
colloid assembly in bulk suspensions.

chiral_colloidal_1

chiral_colloidal_2

chiral_colloidal_3

It is
generally recognized that oil and water do not mix. The physics of such
liquid-liquid phase separation has been known for a century and is seemingly
trivial. Surface tension dictates that droplets assume a spherical shape and
coarsen in time. Furthermore, universally accepted wisdom states that
oil-in-water droplets of finite size can only be stable against coalescence in
the presence of surfactants. We demonstrate that the physics of
membrane-confined 2D colloidal droplets (rafts) profoundly differs from their
3D counterparts. First, membrane droplets do not coarsen with time. Instead
they remain of finite-size indefinitely, and are thermodynamically stable in
absence of any surfactant-like molecules. Second, unlike polydisperse conventional drops,
membrane-mediated interactions lead to self-regulated sizes of 2D droplets,
yielding highly monodisperse structures that can assemble into
higher-order droplet crystals with the ability to self-heal. Third, in contrast
to ubiquitously spherical oil-water droplets, membrane droplets can assume aspherical shapes of essentially arbitrary
complexity.