Colloidal particles trapped at the interface between two immiscible fluids usually display various bidimensional phases depending on the competition between direct and capillary interactions [1]. On the other hand, the physics of colloids confined in a nematic cell has been thoroughly investigated in the recent years. The structures formed by embedded solid spheres strongly depend on the nematic elasticity, the nematic topological defects and the anchoring at interfaces. Only few studies are devoted to colloids at nematic interfaces [2-5] where these interactions combine. Here, we examine the behavior of spherical solid particles at nematic liquid crystal (NLC) interfaces with different geometries. First, we will detail the phase behavior of beads with a strong homeotropic anchoring at an air-NLC interface [6]. Depending on their area density, the nematic thickness and the anchoring, colloids spontanously form various structures (Fig.1). Using optical tweezers, we determine the pair potential and we probe the different roles of capillary and elastic forces. Subsequently, we report the behavior of silica beads at more complex NLC interfaces. By controlling the geometry of the interfaces and the colloids density with microfluidic techniques, several structures have been observed. The competition between anchoring conditions and liquid crystal elasticity, but also the topology of the curved surface is responsible of long-range interactions and the formation of new ordered structures. In particular we examine how the behavior of colloids trapped at the interface of a curved nematic interface modifies the overall director field (Fig.2).

(en cours)

Model systems of charged colloidal particles with non-uniform and fluctuating charge distribution have been designed. They consist of C12EO5 stabilized decane micoremulsions that have been doped with various cationic gemini-type surfactants. The valence of the doping surfactants varies from 1 to 4 and their molecular structure is either linear or branched. The phase diagrams of the doped microemulsions have been established. Close to the emulsification failure temperature, the systems consist of positively charged spherical droplets of about 8nm with Z structural charges, Z varying from 0 to 100. Static light scattering and small angle neutron scattering show that the droplets doped with multivalent and branched surfactants repel each other much less than the ones doped with monovalent or linear multivalent surfactants. To account for these results several parameters were considered: The particles are probably polydispersed in charge as revealed by the observation of both diffusive modes in dynamic light scattering. However, for a given value of Z, the charge polydispersity is essentially determined by the surfactant valence and does not account for the influence of the molecular structure of the surfactant. Counter-ion condensation has been measured by bromide specific electrode and is more pronounced when the doping surfactants are multivalent and branched. This can be attributed to higher local density of charge. Knowing the effective charge of the particles allows calculation of the osmotic compressibility of the solution S(0). When the Debye length is assumed to be determined by the free counter-ions, it means without any open parameter, calculated S(0) is higher than the one obtained experimentally. To recover the experimental S(0) by calculation, an extra screening that reduce the range of the interaction has to be introduced. It varies linearly with the average distance between the patches of charge at the droplet interface and with the molecular structure of the surfactant.

Using particle tracking by optical microscopy, diffusion coefficient of polystyrene and silica particles were measured at the air/water interface. Diffusion is faster at interface than in the bulk water (D2D > D3D) because a significant fraction of the surface area of the particle is submitted to the friction of a low viscosity fluid. To quantitatively account for these experimental results it is crucial to precisely determine the immersion of the particle or equivalently its contact angle. To this purpose we coupled particle tracking and phase shift interferometry, so that the contact angle, the radius and diffusion coefficient are measured on the same particle. Small but significant and reproducible discrepancy between experimental results and theoretical models suggest a role of particle roughness and possible coupling with rotational diffusion.