Design of droplet based microfluidic devices for lab on chip applications

Advances in the area of droplet microfluidics - that can be harnessed to precisely control very small volumes of material in space and time - has the potential to significantly enhance bio-chemical analysis, lab-on-chip technologies and drug discovery. Further, droplet microfluidics holds promise for the development of medical devices that could have an impact in combating deadly diseases such as, for e.g., cancer. The main challenge in developing such devices is the lack of rational design approaches for identifying optimal droplet-based microfluidic architectures based on a desired application. To address this challenge, the current thesis presents key contributions in understanding the basic physics of droplet dynamics through models and development of active and passive techniques to achieve spatio-temporal control of droplets in microfluidic devices.

In the area of modeling, a detailed analysis of existing models is presented. These models are experimentally validated and several enhancements to the existing models are proposed. Using the insights gained from these models and experimental analysis, a theory pertaining to the dynamics of drops at bifurcations and bypasses is proposed. The origins of periodic and chaotic dynamics are uncovered using this theory, experiments and simulations of loop systems. Asymmetrical ladder networks are proposed and theoretically proved to be functionally superior to symmetrical networks. The key findings - such as contraction, expansion, synchronization and flipping in asymmetrical ladder networks - are validated using experiments. Using the insights from the modeling work both active and passive techniques are developed to achieve spatio-temporal control of droplets. In the case of active techniques, several versions of advanced process control strategies are presented. The idea of active resistance modulation for controlling the position and timing of droplets in fluidic networks is proposed. A key finding from this work is that resistance control can help achieve precise sorting and synchronization of droplets at the outlet of a loop device even in cases where the uncontrolled dynamics are chaotic. In the case of passive techniques, we propose an evolutionary algorithm to design ladder networks that can precisely control droplet positions in space and time. A customized GA approach is formulated for designing microfluidic ladder networks. It is shown that the GA approach identifies remarkable and diverse structures for the chosen design problems. Finally, a software is developed - using graph theory and push-pop algorithms - to simulate large scale droplet based microfluidic networks.