High-Speed Droplet Generation in 3D Gas-Liquid Droplet Microfluidic Systems

Generation of liquid droplets within a confined gaseous microflow is a relatively unexplored approach to that of conventional liquid-liquid droplet systems. The creation of uniform particles purely in air (monodisperse aerosols), avoiding cross-contamination of the droplets’ contents by the presence of a secondary liquid is the main advantage. Earlier work has shown that the production of extremely small droplets (15 μm, perhaps submicron) at >100 kHz frequencies, an order of magnitude higher than those of liquid-liquid droplet systems, is possible. The flow regime of gas-liquid droplet microfluidic devices was also characterized to find the conditions necessary for droplet generation in quasi two-dimensional (where all flow channels have the same height), flow-focusing geometries.

Here, we explore and characterize non-planar, ‘three-dimensional’ gas-liquid droplet microfluidic devices to overcome some of the problems that quasi two-dimensional structures have. By ensuring that the carrier phase (liquid) is enveloped by the continuous phase (gas), additional contact of the carrier phase with the flow channel walls is avoided, enhancing throughput performance and system robustness. We use standard 2D planar soft lithography techniques to create 3D architectures instead of custom assemblies needed for each 3D microfluidic device using concentric capillaries. This allows for easier customizability of the flow-focusing geometries to determine optimal architectures.

We have experimentally visualized and comprehensively mapped the flow regime of various 3D gas-liquid droplet microfluidic chip designs via a high-speed camera. We compare the results of the work on quasi-2D chips and discuss how 3D gas-liquid droplet microfluidic systems obtain a higher droplet generation throughput; via a faster transition to the jetting regime by avoiding liquid-wall collisions that delay said transition. Finally, we test the effects of geometry on the flow regime and discuss further optimization to maximize droplet generation.