Research

We are investigating transient gels formed by weakly aggregating emulsions. Transient gels are weak gels that appear to be stable and quiescent for substantial periods of time, but which in due course collapse under the influence of gravity. In the case of a typical emulsion-based system, the result is a cream layer over a layer of clear background fluid. The corresponding process for particles that are heavier than the background gives a sediment layer at the bottom of the container following collapse. We have explored the hypothesis that the formation of fractures or “directed voids” are an essential part of the collapse mechanism.

Conventional microscopy is not very useful for looking deep into an emulsion to explore its structure since there is too much back-scattering. Instead we use confocal microscopy. Even this will not work well with off-the-shelf emulsions since these are opaque, and typically the density difference between the dispersed oil phase and aqueous continuous phase is so large that droplets can also move too much. So our first task has been to construct a novel emulsion system that is both refractive index and buoyancy matched. It’s helpful to have an emulsion in which the droplet size is “monodisperse”. For this reason we create our emulsion using a membrane cross-flow emulsification system, or XME. It turns out that the XME system is essential for creating our novel emulsion due to the properties of the oil.

To get the most from our confocal microscope, which we use in a sideways mode to give good resolution in the direction of gravity, we have created an image processing suite to minimise the impact of the microscope optics as far as possible. Our confocal images are, we believe, the first-ever full three-dimensional images of an emulsion.

We have developed a way of measuring the locations and sizes of the droplets in our three-dimensional images using a technique based on the Euclidean distance map. By stringing together sets of droplet locations over time we can then determine individual droplet trajectories. Droplet locations also allow us to characterise the static structure of the aggregated emulsion, using both conventional Voronoi techniques and a novel “kissing spheres” approach that in our opinion is easier to interpret.

The confocal imaging reveals a rich dynamical behaviour. Single-time stacks sometimes show “dislocations” that are linked to swirls and ripples in the network, some or all of which a attributable to bubbles in the system. In addition, by comparing with the creaming behaviour of the bulk aggregated emulsion, we can see which compositions display the transient gel phenomenon, and search for this using confocal microscopy. In this way we have captured the first-ever microscopic movie of a transient gel entering its collapse phase.

We can also investigate structure at other length and time scales. Magnetic resonance imaging lets us look for structure at length scales of tens of microns and in samples that are not necessarily transparent. In addition computer simulations allow us to explore what happens during the formation of the aggregated network, in the expectation that this initial structure will influence the evolution of the system over time.

We do not pretend to have all the answers, but our work offers some useful insights.

View our Conclusions
View our Publications