Passive and Deterministic Droplet Microfluidics for Single Object Encapsulation and Pairing

The single-cell study, which dissects cell-cell heterogeneity from bulk populations, is of growing attention and importance in biological studies and clinical practices. At the core of the single-cell analysis is the efficient cell isolation and construction of single-cell assays. Among different platforms, droplet-based microfluidics arises as one of the most popular ones, However, like all other tools, the isolation of single cells into each compartment (i.e., droplet) is confronted with randomness, where the cell occupancy in each compartment is non-deterministic but follows Poisson statistics (Poisson limit). The Poisson limit is currently an inevitable obstacle in droplet microfluidics, hindering a large number of droplet-based single-cell assays.

Attempts have been made to overcome the Poisson limit during single-cell encapsulation into droplets. Active approaches requiring droplet detection, actuation, and synchronization, is complex, expensive, and limited by throughput (label-free). On the other hand, the passive strategy is automatic and simpler, but the current methods lack control and robustness on the single-cell encapsulation outcomes. A passive yet deterministic, simple, and robust approach for single-cell encapsulation is missing. Active or passive, there is no droplet microfluidics platform for label-free and ultra-high throughput single-cell encapsulation, nor for distinguishing droplets containing one cell from droplets with multiple cells. This thesis targets on these issues.

In this thesis, we first looked into details of a new type of droplet instability and studied the corresponding new breakup mechanism analytically, experimentally, and numerically. This part is presented in Chapter 2. Chapters 3, 4, and 5 each present a strategy toward passive single-cell deterministic encapsulation. In Chapter 3, we used a post-encapsulation secondary breakup-mediated droplet sorting strategy, based on the phenomenon of cell-triggered (droplet) splitting (CTS). The droplet instability from Chapter 2 is the cause of the CTS, which is unique and scalable with a simple geometrical rule. In Chapter 4, we developed a post-encapsulation direct droplet sorting strategy, based on the single-cell differentiated droplet trajectory (SCDT). We demonstrated a novel passive mechanism to induce a change of droplet trajectory based on the number of single cells it contains. Empty droplets, single-cell droplets, and multi-cell droplets can be equally separated. In Chapter 5, towards the deterministic rigid particle encapsulation, we developed a particle-triggered droplet generation technology (K-blocking), enabled by a novel fabrication strategy. This is a passive on-demand droplet generator. Finally, in Chapter 6, we developed a waiting-concept-aided deterministic droplet pairing and merging device (In-flow merging). We demonstrated 100% one-to-one merging for one droplet type using this mechanism, particularly suited for low-input droplet sample