Fractionation by shape in deterministic lateral displacement microfluidic devices

TL;DR

This study explores how particles of different shapes and sizes migrate in a gravity-driven deterministic lateral displacement device, revealing shape and size dependencies that can improve particle separation techniques.

Contribution

It demonstrates the dependence of critical angle on particle shape and size, and introduces a potential universal relationship using inscribed sphere diameter for device optimization.

Findings

Critical angle varies with particle size and shape.

Using inscribed sphere diameter makes the critical angle shape-independent.

Linear relationship between critical angle and inscribed sphere diameter.

Abstract

We investigate the migration of particles of different geometrical shapes and sizes in a scaled-up model of a gravity-driven deterministic lateral displacement (g-DLD) device. Specifically, particles move through a square array of cylindrical posts as they settle under the action of gravity. We performed experiments that cover a broad range of orientations of the driving force (gravity) with respect to the columns (or rows) in the square array of posts. We observe that as the forcing angle increases particles initially locked to move parallel to the columns in the array begin to move across the columns of obstacles and migrate at angles different from zero. We measure the probability that a particle would move across a column of obstacles, and define the critical angle {\theta}c as the forcing angle at which this probability is 1/2. We show that critical angle depends both on particle size and shape, thus enabling both size- and shape-based separations. Finally, we show that using the diameter of the inscribed sphere as the characteristic size of the particles the corresponding critical angle becomes independent of particle shape and the relationship between them is linear. This linear and possibly universal behavior of the critical angle as a function of the diameter of the inscribed sphere could provide guidance in the design and optimization of g-DLD devices used for shape-based separation.