Separation of deformable particles in deterministic lateral displacement devices

Abstract

Using numerical simulations, we study the separation of deformable bodies, such as capsules, vesicles, and cells, in deterministic lateral displacement devices, also known as bump arrays. These arrays comprise regular rows of obstacles such as micropillars whose arrangements are shifted between adjacent rows by a fixed amount. We show that, in addition to the zigzag and laterally displaced trajectories that have been observed experimentally, there exists a third type of trajectory which we call dispersive, characterized by seemingly random bumpings off the micropillars. These dispersive trajectories are observed only for large and rigid particles whose diameters are approximately more than half the gap size between micropillars and whose stiffness exceeds approximately 500 MPa. We then map out the regions in phase space, spanned by the row shift, row separation, particle diameter, and particle deformability, in which the different types of trajectories are expected. We also show that, in this phase space, it is possible to transition from zigzag to dispersive trajectories, bypassing lateral displacement. Experimentally, this is undesirable because it limits the ability of the device to sort particles according to size. Finally, we discuss how our numerical simulations may be of use in device prototyping and optimization.

Effects of the particle deformability on the critical separation diameter in the deterministic lateral displacement device

“…the case of G s = 0.01 in figure 5c). As mentioned in the introduction, Beech et al (2012) observed that the blood cell was more prone to the zigzag model, as its deformability increased, whereas the numerical results of Quek et al (2011) showed the opposite tendency. This contradiction can now be resolved from our simulations, since the particle deformability of Beech et al (2012) was notably large whereas that of Quek et al (2011) was relatively small.…”

“…Quek et al (2011) employed the immersed boundary method to investigate the effects of various control parameters on the separation of deformable particles in the DLD device, and observed that with the increase of particle deformability, the zigzag mode could be turned into the lateral-displacement mode, which is qualitatively opposite to the experimental observation of Beech et al (2012). Thus, further studies need to be carried out to understand the effects of particle deformability, which is the main aim of the present study.…”

“…∂u/∂t + U 0 ∂u/∂x = 0) is imposed at the outlet. Following Quek et al (2011), the periodic boundary condition is introduced in the lateral direction, which is suitable for the realistic DLD device with a large number of post rows. Since we use the inlet and outlet boundary conditions in the streamwise direction, the number of the posts (or the size of the computational domain) needs to be large enough to capture the motion mode of the particle.…”

Optimizing Polymer Lab-on-Chip Platforms for Ultrasonic Manipulation: Influence of the Substrate

“…Combining microfluidic systems with different readout techniques has produced powerful tools in the emerging area of the on-chip controlled handling and monitoring of biological and synthetic microparticles and nanoparticles, with especially great impact on biotechnology and environmental topics, such as hydrodynamic forces [1][2][3][4][5][6][7][8][9][10][11][12][13][14][15][16][17][18][19][20], optical methods [21,22], electric fields and dielectrophoresis [23][24][25][26][27] or acoustic fields , either isolated or combined with other fields.…”

Inertia and scaling in deterministic lateral displacement

“…[1][2][3][4][5][6][7][8][9][10] It has, in fact, been successfully used in a number of applications, especially for the separation of biological samples. [11][12][13][14][15][16][17][18] Although in most cases DLD is described (and investigated) as a size-based separation technique, it could also separate particles based on shape and flexibility 19,20 and may contribute to the growing field of chiral separation. 21,22 We have shown that a simple model 23,24 that describes the motion of a suspended sphere around an individual obstacle in the DLD array (which we refer to as a particle-obstacle collision) predicts the separation capability of the system.…”