Hydrodynamic Chromatography with Deterministic Lateral Displacement Effect

Abstract

Hydrodynamic chromatography (HDC) is a flow-driven passive method for separating micrometric/nanometric particles based on the interaction between a nonuniform velocity profile and Brownian diffusion, which causes particles of different size to migrate at different average velocity throughout the separation column. Despite its conceptual simplicity and relative ease of implementation, HDC remains to date an underutilized technique in view of the lengthy channels and large operational times required. In the search for optimal geometries enhancing separation efficiency, micro-Pillar Array Columns (μPACs), constituted by a doubly periodic obstacle lattice aligned with the direction of the flow, have been successfully proposed and tested. The aim of this article is to show that a further improvement of HDC efficiency in μPACs is possible by enforcing a symmetry breakup, where the lattice is misaligned by an angle θ_l with respect to the flow direction. The mismatch between the flow direction and the lattice axes triggers a new separation mechanism, referred to as Deterministic Lateral Displacement (DLD), which causes particles of different size to migrate along different directions through the lattice. So far, DLD has been enforced exclusively in continuous separations run under steady-state conditions.. If an unsteady (chromatographic) operating mode in a slanted μPACs is enforced, differences in migration velocities and migration angles act simultaneously as two independent mechanisms. Theoretical/numerical evidence is provided, showing that the synergy between the two separation drives can shorten device lengths and analysis times by a factor of 10 or even higher (depending on the analytical target) when compared to plain-HDC. The results presented are based on an advection-diffusion template enforcing the classical excluded-volume model to account for particle–wall interactions, an approach previously validated against experimental data by different research groups, both in standard μPACs-HDC and in continuous DLD devices. Numerical results of the average particle migration angle and velocity magnitude are obtained by two independent (Eulerian and Lagrangian) computational approaches. A case study of geometry is used throughout to illustrate the concrete implementation of the method for a multidispersed mixture of particles of five nominal diameters ranging from 1 to 1.6 μm.