Beyond Laminar Mixing: the Advantages of Combining Mixing Approaches in Microfluidic Chips

The effective mixing of two or more fluids is a key target for virtually all microfluidic-assisted processes. However, making this action rapid is generally difficult because of the inherently laminar character of the flow in small channels (low Reynolds / high Peclet numbers). The most common approaches (passive micromixers) are the induction of chaotic features (e.g. staggered herringbone chips) and the enhancement of liquid-liquid interfaces (e.g. split-and-recombine chips). This study focuses on a micromixer design that combines a split-and-recombine approach (effective already at low Reynolds numbers) with slanted-groove connectors (boosting chaotic dynamics at high Re ). Both experiments (epifluorescence and confocal microscopy using low-diffusion fluorescent probes) on 3D printed chips, and Computational Fluid Dynamics (CFD) simulations (particle tracing) show that this combination is particularly effective at intermediate to high Re . For example, already after a first mixing unit, at Re = 10 the mixing efficiency is 45-60% (depending on the chip junction) vs. ≈30% for a staggered herringbone chip with comparable length, and further increases at higher Re . At low Re , where the chip predominantly acts as a split-and-recombine mixer, we have found that increasing the amount of liquid-liquid interface at the junction point is a particularly useful approach; this increases the mixing efficiency from e.g. 35-40% with a simple Y junction, to 50-55% with 2D flow-focusing. In terms of chip design, our results therefore show that the combination of flow-focused split-and-recombine (low-Reynolds) and chaotic flow (high-Reynolds) is a particularly effective and additive combination to enhance mixing behaviour.