Enhancing particle focusing: a comparative experimental study of modified square wave and square wave microchannels in lift and Dean vortex regimes

Abstract

AbstractSerpentine microchannels are known for their effective particle focusing through Dean flow-induced rotational effects, which are used in compact designs for size-dependent focusing in medical diagnostics. This study explores square serpentine microchannels, a geometry that has recently gained prominence in inertial microfluidics, and presents a modification of square wave microchannels for improved particle separation and focusing. The proposed modification incorporates an additional U-shaped unit to convert the square wave microchannel into a non-axisymmetric structure, which enhances the Dean flow and consequently increases the Dean drag force. Extensive experiments were conducted covering a wide range of Reynolds numbers and particle sizes (2.45 µm to 12 µm). The particle concentration capability and streak position dynamics of the two structures were compared in detail. The results indicate that the modified square-wave microchannel exhibits efficient particle separation in the lower part of the Dean vortex-dominated regime. With increasing Reynolds number, the particles are successively focused into two streaks in the lift force-dominated regime and into a single streak in the Dean vortex-dominated regime, in this modified square wave geometry. These streaks have a low standard deviation around a mean value. In the Dean vortex-dominated regime, the location of the particle stream is highly dependent on the particle size, which allows good particle separation. Particle focusing occurs at lower Reynolds numbers in both the lift-dominated and lift/Dean drag-dominated regions than in the square wave microchannel. The innovative serpentine channel is particularly useful for the Dean drag-dominated regime and introduces a unique asymmetry that affects the particle focusing dynamics. The proposed device offers significant advantages in terms of efficiency, parallelization, footprint, and throughput over existing geometries.