Colloids at interfaces: Pinned down.

In 1903 Walter Ramsden found that an air bubble placed in a liquid suspension of microscopic particles “can be seen to pick up the particles ... and to retain them obstinately ...”1 Ramsden had discovered the strong affinity of colloidal particles to fluid interfaces, an affinity that a century later can be harnessed to make microstructured materials such as Pickering emulsions, colloidosomes and bijels2. The production of these materials relies not only on the particles obstinately sticking to the interface, but also on their ability to move about and self-assemble. In this regard, Maurizio Nobili and colleagues3 show in Nature Materials that the motion of a colloidal particle confined to lie at an air/water interface is, surprisingly, slower than the motion of the particle in the bulk of the fluid. To understand this finding, one ought to consider the classical picture of how a colloidal particle sticks to an air/water interface4 (Fig. 1a). If the particle is a smooth sphere, and small enough so that its weight can be neglected, its equilibrium position with respect to the fluid interface can be determined by minimizing the sum of three interfacial energies (those corresponding to the water/air, water/ particle, and air/particle interfaces). For typical values of the interfacial tensions, a micrometre-sized particle that adsorbs to the interface lowers the total interfacial energy by 103–106 times the thermal energy. This energetic argument explains why the interface can so ‘obstinately’ retain the particles. In the lowest-energy configuration, the interface remains flat, and the top of the particle protrudes into the upper phase by an amount determined by the contact angle (as described by Young’s equation). However, the equilibrium state is not a static situation: a colloidal particle is always in motion owing to the constant and random bombardment of solvent molecules. As Einstein showed, the diffusion coefficient that characterizes this Brownian motion in the bulk varies inversely with the drag force on the particle, which is proportional to the fluid’s dynamic viscosity. Hence, one expects that the diffusion coefficient of a particle sitting at the air/liquid interface should increase with the contact angle, because the larger its magnitude the more the particle protrudes into the less viscous air phase. The same prediction follows from more rigorous hydrodynamic arguments5. Yet by measuring the particle’s Brownian motion in the plane of the interface (and avoiding any contamination that might affect its properties), Nobili and co-authors found exactly the opposite trend: as they increased the contact angle (by using progressively more hydrophobic particles), the particles diffused more slowly. They observed this same trend for a variety of different particles, which suggests that the phenomenon originates from some common feature of the particles. The authors propose that this common feature is nanoscale heterogeneity. At some small scale, all solid surfaces are irregular, because of the presence of surface functional groups or topographical features (or both). These ‘defects’ can pin a contact line. If the pinning is much stronger than the thermal energy, the local contact angle might fluctuate as thermal capillary waves lap at the boundary of the particle; if the pinning is weak, the position of the contact line might hop randomly from defect to defect, driven by the thermal energy. In either case, the fluctuation–dissipation theorem predicts that these fluctuations must couple to a dissipative or drag force. Such additional dissipation adds to the usual drag force on a particle in a viscous fluid, which leads to a reduced diffusion coefficient. At present, this explanation is supported only by indirect evidence. Nobili and colleagues show that the length scale between defects, determined by fitting two different pinning models to the data, is on the order of a nanometre, and hence comparable to the expected distance between functional groups on the particle’s surface. But such small defects have never been seen directly, and there is no evidence that the functional groups are capable of pinning the contact line. Nevertheless, the results are consistent with a growing body of work showing — in the same, indirect way — that nanoscale heterogeneities do exist in colloids and are important for understanding their behaviour at fluid interfaces. In fact, recent experiments have shown that the attractive interactions between particles at interfaces6, as well as the relaxation rates of particles that breach an interface7,8, appear to be governed by surface defects. And because colloids typically have