My research projects are mainly related to colloid science problems that can be investigated with optical tweewers, and nucleation:

Perturbing and probing baths of colloids with optical tweezers

Optical tweezers are a great tool in microrheology to both perturb locally a system and to measure its response. In optical tweezers, laser beams are focalized on particles whose refractive index differ from the fluid, thus resulting in a force on the particles. We use optical tweezers to study colloidal solutions. Baths of colloidal particles are widely used as models of atomic systems to study phase transitions and nucleation, and the local and controlled changes that optical tweezers can apply onto a system provides a powerful tool to study microscopic phenomena occuring in colloidal baths.

In the first movie below, the sample is a bath of fluorescent PMMA colloids that are density- and index-matched with the organic solvent that surrounds them. A polymer has also been dissolved to introduce a depletion interaction. The system is perturbed by adding bigger colloids that are not index-matched. These can be trapped and heated with optical tweezers. By controlling the laser power we can create a local aggregation of the PMMA colloids around the trapped particles because of thermophoresis. The second movie shows what happens when the sample starts being moved at a constant speed with the colloid still trapped. At this large velocity the shell of a few layers of small colloids that forms the aggregate is wiped by the flow.

We expect that such aggregates could be used to create locally controlled phase transitions to study phenomena such as nucleation.

This video shows a trapped, big colloid (central void) in a bath of small PMMA colloids that do not interact with the laser beam. The laser power of the tweezers switches between two values (low and high) alternatively. At high power, an aggregate of PMMA colloids forms around the trapped particle.

Here, the system is the same as before, except that laser power is constant at a high value, and that the whole sample starts being moved at a constant speed to create a flow of the fluid around the trapped particle. The flow has a high enough value so that the aggregate is entirely destroyed. At lower flow velocity, the lower layers of the aggregate would stay around the particle, which results in a strong nonlinearity of the drag force as the velocity decreases.

BCC aggregate
Pentagonal aggregate

(a) Aggregate made of crystal domains with BCC structure. (b) By tuning the sizes of the central and fluorescent colloids, we can constrain the system to form unusual structures.

Hydrodynamic interaction and synchronization between active colloids

The main motivation in this project is to understand here the hydrodynamic synchronization of cilia and flagella. Motile cilia and flagella are thin biological filaments that are present at the surface of some cells. They bend in a cyclic pattern to generate a flow of the fluid that surrounds them. It is known that assemblies of cilia coordinate their beats and show some degree of synchrony. This work investigates the role of the hydrodynamic interaction on the coordination properties of such oscillators. The knowledge gained in this field is relevant to the development of artificial swimmers as well as the diagnosis of diseases that involve the loss of motility of the active cilia.

SEM picture of Bronchiolar epithelium

SEM picture of Bronchiolar epithelium showing cilia. Each cilium is typically 3 to 5 µm long.

Optical tweezers can help! We model an oscillating cilium by a colloid driven by optical tweezers in a cyclic pattern. This model retains most of the physics of cilia (far field hydrodynamic interactions at low Re number and presence of thermal fluctuation). Since all the complex dynamics of bending of the cilia is coarse-grained into spherical active colloids, it is possible to identify the key parameters that control synchronization and the experiments can be complemented by Brownian Dynamics simulations and analytical calculations.

Geometric switch

Example of model oscillator. In this animation, an oscillating cilium is represented by a colloid moving between two traps that are switched on and off in turn. By putting several oscillators like this one close to each other, it becomes possible to study their synchronization.

Using such systems, we have been able to show that hydrodynamic coupling very often leads to strong synchronization. For two oscillators, we have studied how the type of the drive or the shape of the driving potentials change the state and strength of synchronization. For systems of more than two oscillators, complex dynamic phase patterns are seen, and we have started to characterize the emergence of metachronal waves. A summary of our results can be found in this review article.

Synchronization results

A few results. (a) Synchronization map of two oscillators like in the animated cartoon above. The parameters that control the synchronization are the curvature of the driving force, c, and the temperature, ξ (in an adimensional form). The temperature is relevant in this diagram, as thermal fluctuations of the colloids can destroy a weakly synchronized state. Legend: P: in phase; AP: in antiphase; NS: not synchronized. Real systems such a the unicellular biflagellated alga Chlamydomonas reinhardtii can be represented on this diagram to predict their synchronization state. (b) Another type of oscillator we have studied is the rotor. Here, each oscillator is a colloid driven on a circular path. The top sketch shows a chain of six rotors. When set into motion, the system can synchronize in a metachronal wave (diagonal wave fronts in the bottom phase plot).

Many configurations of oscillators can be studied, including a set of five rotors arranged as olympic rings. The rotors can synchronize, as shown in this movie.

Until recently, all the experiments were performed in a Newtonian fluid. The biological fluids in which cilia move are however often vicoelastic which is the object of current investigations.

Probing the metastability of liquids with acoustic cavitation

This work is being done in the group of Frédéric Caupin.

Cavitation is the process of creating, from a metastable liquid, a stable gas phase that will grow indefinitely. This happens for example when the pressure of a liquid is lowered below a certain value. Good estimates of nucleation thresholds are required by many applications or to prevent disasters (filling/emptying of nanopores, climate models, spill accidents...).

Phase diagram

Phase diagram of a common liquid. The blue lines are the boundaries of the stable regions: solid, liquid and vapor. The ultimate pressure that can be reached without cavitating is indicated by the red dashed line below the stable limit. Condensation also diplays metastability, which is shown by the upper dashed line.

Nucleation is usually described by classical nucleation theory (CNT), but it often leads to inaccurate nucleation thresholds. CNT relies on the calculation of the work required to create a nucleus of a given size. This work is highly dependent on the surface tension associated to the liquid-gas interface that forms the nucleus. At the nanoscale, the interface is however not very well defined, and surface tension is therefore not well defined either. In this project, I was measuring cavitation thresholds in order to test models based on CNT, but with radius-dependent formulas for the surface tension. We have shown that the most common correction to the surface tension (Tolman equation) does not describe properly nucleation data from experiments.

Tolman equation

Possible dependency of the surface tension σ with the radius of a bubble (-) or of a droplet (+). The parameter δ is the Tolman length.

Experimentally, we lower the pressure of a liquid to negative values with a piezo-electric transducer (acoustic cavitation) and pressures are measured by monitoring changes in the refractive index of the liquid (hence in the density) with an optical fiber that allows to measure the reflective coefficient at the glass/liquid interface of the fiber. The discrepancy between the nucleation thresholds from the experiments and from the CNT predictions gives an insight into the shape of the R-dependent surface tension.

Last update: 2017-03-10