We investigate the separation of drops in force-driven deterministic lateral displacement (f-DLD), a promising high-throughput continuous separation method in microfluidics. We perform scaled-up macroscopic experiments in which drops settle through a square array of cylindrical obstacles. These experiments demonstrate the separation capabilities-and provide insight for the design-of f-DLD for drops of multiple sizes, including drops that are larger than the gaps between cylinders and exhibit substantial deformation as they move through the array. We show that for any orientation of the driving force relative to the array of obstacles, the trajectories of the drops follow selected locking directions in the lattice. We also found that a simple collision model accurately describes the average migration angles of the drops for the entire range of sizes investigated here, and for all forcing directions. In addition, we found a difference of approximately 20° between the critical angles at which the smallest and largest drops first move across a line of obstacles (column) in the array, a promising result in terms of potential size resolution of this method. Finally, we demonstrate that a single line of cylindrical obstacles rotated with respect to the driving force is capable of performing binary separations. The critical angles obtained in such single line experiments, moreover, agree with those obtained using the full array, thus validating the assumption in which the trajectory (and average migration angle) of the drops is calculated from individual obstacle-drop collisions.

The ability to separate and analyze chemical species with high resolution, sensitivity, and throughput is central to the development of microfluidics systems. Deterministic lateral displacement (DLD) is a continuous separation method based on the transport of species through an array of obstacles. In the case of force-driven DLD (f-DLD), size-based separation can be modelled effectively using a simple particle-obstacle collision model. We use a macroscopic model to study f-DLD and demonstrate, via a simple scaling, that the method is indeed predominantly a sizebased phenomenon at low Reynolds numbers. More importantly, we demonstrate that inertia effects provide the additional capability to separate same size particles but of different densities and could enhance separation at high throughput conditions. We also show that a direct conversion of macroscopic results to microfluidic settings is possible with a simple scaling based on the size of the obstacles that results in a universal curve. V C 2013 AIP Publishing LLC.