Electronics, Pneumatics and Computer Vision for Advancing Droplet-based Microfluidic Automation

Abstract

This dissertation explores the subdomain of droplet-based microfluidics flow control, a field of study to precisely manipulate multiphase flow segmentation at the microscale. The focal point of this research is to advance the automation of droplet-based microfluidic systems through the synergistic integration of electronics, pneumatics, and computer vision. By addressing the limita-tions inherent in existing methodologies, this work lays the groundwork for the development of innovative applications. Chapter 1 introduces droplet microfluidics, highlighting its development from stable emulsions to a versatile tool in scientific research. The chapter reviews key advancements in droplet manipu-lation and analysis, enabling high-throughput screening and precise control. It discusses passive and active droplet control and the integration of optical detection methods. Innovations in fluidic control and unique droplet junction designs have been addressed. Despite these advancements, challenges remain integrating solid-phase techniques and enhancing the accuracy and efficiency of droplet-based systems for broader use. Identifying the key challenges, this chapter provides the outline for the rest of the dissertation where different power-based control systems have been explored in the context of droplet microfluidics. Chapter 2 focuses on the design and fabrication of microfluidic devices using direct 3D printing and polydimethylsiloxane (PDMS) with 3D-printed templates. The reproducibility of both meth-ods has been validated through the development of various custom microenvironments. The chapter provides a detailed discussion on the curing chemistry of PDMS on 3D templates, fol-lowed by the development, characterization, and application of several microfluidic flow control systems for single and multiple, laminar, and segmented flows. Chapter 3 explores the automation of microfluidic droplet generation and manipulation through computer programming, utilizing LabVIEW and Arduino to orchestrate a ‘normally closed’ valve system. A multi-step microfluidic automation has been proposed, designed, and validated on the system developed in the previous chapter. This chapter also introduces a serial communication-based valve control system capable of executing multiple operations simultaneously and a high-resolution pressure sensor for detecting rapid pressure changes, to be integrated for device func-tionality and control. Chapter 4 transitions to a pneumatic automation approach, employing in-house designed, 3D-printed and assembled logic gates for droplet-based manipulation. Characterization of the fun-damental pneumatic logic gates is followed by their applications in controlling single and multiple phase aqueous in oil droplet generation. After demonstrating how the entire droplet generation process can be completely automated without electronic power, the development of a pneumatic multiplexer illustrates a significant move towards minimizing electronic dependency, showcasing the feasibility of a fully pneumatic-controlled microfluidic system. Chapter 5 introduces computer vision – integrated with serial communication - as a transforma-tive tool for real-time feedback and control of droplet generation. By analyzing microscopic data through Python programming, the system dynamically adjusts valve actuation, optimizing droplet formation and demonstrating a leap towards precision and efficiency in microfluidic automation. In Chapter 6, the dissertation concludes by contemplating future research directions, emphasiz-ing the potential of these technologies to foster fully autonomous, electronic-less microfluidic devices for a wide array of applications.