Droplet microfluidics is a high-throughput platform capable of accelerating the screening and synthesis of biological and chemical systems. However, significant challenges to microfluidic design and fabrication limit its broad use. In this dissertation, I overview and present potential solutions to challenges in droplet microfluidic fabrication and design.
First, I present a method for low-cost rapid prototyping of complex droplet microfluidic devices. By combining desktop micromilling and electrode integration with conductive ink, thermoplastic microfluidic devices can be produced with features as small as 75 microns that can be connected to external sensors and actuators. Such devices can be designed, fabricated, and tested within a day and are shelf stable for months. Next, I developed a droplet microfluidic component library using micromilling and conductive ink electrodes. This library is high-throughput, biocompatible, and consists of components for droplet generation, anchoring, reinjection, coalescence, picoinjection, capacitance sensing, fluorescent sensing, and sorting. These components were combined in complex workflows, specifically, in the development of multi-color droplet pixel arrays. Finally, a series of machine learning based design automation tools for droplet microfluidics were created. These tools are capable of predicting the performance and automating the design of single emulsion and double emulsion droplet generators across any fluid combination. Furthermore, two quality metrics were developed, versatility and flow stability, that provide important context on the behaviors of the suggested designs. These tools are the first of their kind in microfluidics, and can play an important role in shifting droplet microfluidic design away from the manual and iterative process it is today.
These advancements in droplet microfluidic design and fabrication can set the basis to rethink the microfluidic development cycle. Predictable and reproducible design and fabrication of sophisticated droplet microfluidic devices would provide a next-generation automation platform for the biological and chemical sciences, running experiments orders of magnitude faster and more sensitive than current methods. / 2025-01-16T00:00:00Z
| Identifer | oai:union.ndltd.org:bu.edu/oai:open.bu.edu:2144/45469 |
| Date | 17 January 2023 |
| Creators | McIntyre, David Patrick |
| Contributors | Densmore, Douglas M. |
| Source Sets | Boston University |
| Language | en_US |
| Detected Language | English |
| Type | Thesis/Dissertation |
| Rights | Attribution 4.0 International, http://creativecommons.org/licenses/by/4.0/ |
Page generated in 0.0025 seconds