Particle focusing in planar geometries is essentially required in order to develop cost-effective lab-on-a-chips, such as cell counting and point-of-care (POC) devices. In this study, a novel method for sheathless particle focusing, called "Elasto-Inertial Particle Focusing", was demonstrated in a straight microchannel. The particles were notably aligned along the centerline of the straight channel under a pressure-driven flow without any additional external force or apparatus after the addition of an elasticity enhancer: PEO (poly(ethylene oxide)) (∼O(100) ppm). As theoretically predicted (elasticity number: El≈O(100)), multiple equilibrium positions (centerline and corners) were observed for the viscoelastic flow without inertia, whereas three-dimensional particle focusing only occurred when neither the elasticity nor the inertia was negligible. Therefore, the three-dimensional particle focusing mechanism was attributed to the synergetic combination of the elasticity and the inertia (elasticity number: El≈O(1-10)). Furthermore, from the size dependence of the elastic force upon particles, we demonstrated that a mixture of 5.9 and 2.4 µm particles was separated at the exit of the channel in viscoelastic flows. We expect that this method can contribute to develop the miniaturized flow cytometry and microdevices for cell and particle manipulation.

No abstract

Dynamic self-assembly of droplets, regular structure formation of moving deformable objects in a confinement environment is a challenging problem in nonlinear dynamics and engineering patterned structure. In the current work, we investigated how the local kinematic history affects the dynamic self-assembly of picoliter-sized droplets near the expansion regions in microfluidic devices. The local kinematic history was controlled by the shape of the expansion region and characterized using computational fluid dynamics. Sizecontrolled aqueous droplets in light mineral oil were continuously generated at T-junction microchannel and transported toward the expansion region. The fast dynamics of the droplets was tracked using high-speed video microscopy. We found three types of dynamic droplet arrays: 1D, 2D zigzag, and irregular. The orderdisorder transition was associated not only with the droplet size, but also with the controlled local kinematic history, which results in the transient deformation of droplet and droplet-droplet interactions. The present results provide us with insight into the dynamic self-assembly of droplets and could be a useful guide for practical applications of droplet-based microfluidics.