Hydrodynamic microfluidic platforms have been proven to be useful and versatile for precisely sorting particles/cells based on their physicochemical properties. In this study, we demonstrate that a simple lattice-shaped microfluidic pattern can work as a virtual sieve for size-dependent continuous particle sorting. The lattice is composed of two types of microchannels ("main channels" and "separation channels"). These channels cross each other in a perpendicular fashion, and are slanted against the macroscopic flow direction. The difference in the densities of these channels generates an asymmetric flow distribution at each intersection. Smaller particles flow along the streamline, whereas larger particles are filtered and gradually separated from the stream, resulting in continuous particle sorting. We successfully sorted microparticles based on size with high accuracy, and clearly showed that geometric parameters, including the channel density and the slant angle, critically affect the sorting behaviors of particles. Leukocyte sorting and monocyte purification directly from diluted blood samples have been demonstrated as biomedical applications. The presented system for particle/cell sorting would become a simple but versatile unit operation in microfluidic apparatus for chemical/biological experiments and manipulations.

We herein propose a new hydrodynamic mechanism of particle separation using dual-depth, lattice-patterned asymmetric microchannel networks. This mechanism utilizes three-dimensional (3D) laminar flow profiles formed at intersections of lattice channels. Large particles, primarily flowing near the bottom surface, frequently enter the shallower channels (separation channels), whereas smaller particles flowing near the microchannel ceiling primarily flow along the deeper channels (main channels). Consequently, size-based continuous particle separation was achieved in the lateral direction in the lattice area. We confirmed that the depth of the main channel was a critical factor dominating the particle separation efficiencies, and the combination of 15-μm-deep separation channels and 40-μm-deep main channels demonstrated the good separation ability for 3–10-μm particles. We prepared several types of microchannels and successfully tuned the particle separation size. Furthermore, the input position of the particle suspension was controlled by adjusting the input flow rates and/or using a Y-shaped inlet connector that resulted in a significant improvement in the separation precision. The presented concept is a good example of a new type of microfluidic particle separation mechanism using 3D flows and may potentially be applicable to the sorting of various types of micrometer-sized objects, including living cells and synthetic microparticles.

No abstract