Abstract:
This thesis aims to study the dynamics of colloidal particles in a microfluidic environment. This topic is of fundamental interest because controlled microfluidic experiments can provide insights into the assembly of individual colloidal building blocks, e.g. Janus particles, into complex hierarchical structures. Aggregation behaviours specifically enabled by microfluidics, and possible applications arising from these, are also of interest. Our studies have primarily focused on understanding the dynamics of hydrophobic silica colloids, and of plain and amphiphilic Janus particles.
Firstly, we have investigated the diffusion behaviour of hydrophobic silica colloids using a combination of free diffusion, microfluidics, and optical tweezers methods. The aim was to detect and quantify the intrinsic slip phenomenon based on the dynamics of particle diffusion, which is expected to vary in the presence of a slip boundary condition. It was hypothesized that a particle experiencing slip would diffuse faster and have a smaller hydrodynamic radius. Silica colloids were coated with OTS (octadecyltrichlorosilane) monolayers for the diffusion studies. We observed an increase in the hydrodynamic radii of these particles by more than two times the OTS molecule length (~ 2.62 nm), even for coating conditions which produced a monolayer verified using ellipsometry, and for which no polymerization was observed using SEM. The increased hydrodynamic radius indicated that a slip boundary condition was not detected on the particle surface. Overall, this work established a method to detect and quantify the effect of chemical modification at the fluid-solid interface on a diffusing particle.
Secondly, different microfluidic chip configurations, namely, co-flow focusing, sheath-flow focusing, and particle trap-release method, were fabricated and tested with the aim of controlling Janus particle interaction and assembly. Co-flow focusing is a relatively simple method, but it was observed to be inefficient in focusing Janus particles into tight streamlines, leading to relatively few interaction events. The hydrodynamic trap and release method offered precise control over Janus particles assembly. However, the frequency of the particle interactions was limited due to the challenging nature of controlling the simultaneous release of Janus particles from the traps. Sheath-flow focusing provided an excellent balance between co-flow focusing and particle trap and release. Moreover, sheath-flow focusing offered a good control of the focused particle stream, thereby increasing Janus particle interactions.
Finally, the dynamics of individual Janus particles and their interactions were studied inside a microchannel using sheath-flow focusing. The angular rotations of Janus particles were used to conduct a comparative study between the dynamics of individual and interacting plain and amphiphilic Janus particles near the centre and the wall of the channel. It was observed that the average angular velocity of individual amphiphilic Janus particles may be slightly lower than for plain Janus particles, although not with good statistical significance. To the best of our knowledge, our method is the first to directly measure colloidal Janus particle rotation rates in rectangular microchannels, showing slightly lower angular velocity than predicted in pure shear.
Moreover, a reliable method was developed for the clear observation of the individual and interacting Janus particle orientations in a continuous flow. Changes in the orientations of amphiphilic Janus particles with close initial trajectories have been measured. Observed variations in particle orientation could be due to the possible hydrophobic-hydrophobic interactions; however, the evidence is not conclusive. Similarly, our findings have shown that the relative orientations of Janus particles are not affected by surface interactions in a static environment. Additionally, it was observed that the Janus particle orientations in dimers moving near the microchannel wall and the microchannel centre are not affected by high and low shear rates.