(257j) Numerical Simulation Study of the Deterministic Vector Chromatography of Rigid and Flexible Particles Over Slanted Open Cavities

Authors: 
Bernate, J. A., Johns Hopkins University
Yang, M., Stanford University
Zhao, H., Stanford University
Risbud, S. R., Johns Hopkins University
Paul, C. D., University of Arkansas
Dallas, M., Johns Hopkins University
Konstantopoulos, K., Johns Hopkins University
Drazer, G., Rutgers University
Shaqfeh, E. S. G., Stanford University



Planar microfluidic platforms for vector chromatography, in which different species fan out in different directions and can be continuously sorted, are particularly promising for the high throughput separation of multicomponent mixtures. We carry out a computational study of the vector separation of dilute suspensions of rigid and flexible particles transported by a pressure-driven flow over an array of slanted open cavities. The numerical scheme is based on a Stokes flow boundary integral equation method. The simulations are performed in a periodic system without lateral confinement, relevant to microfluidic devices with much larger width than the dimensions of the open cavities and the ridges that create them.  We study the deflection of rigid spherical particles, of  flexible capsules as a model of  white and red blood cells,  and of rigid discoidal particles as a model of platelets. We characterize the deflection of different particles as a function of their size, shape, shear elasticity, their release position, and the geometric parameters of the channel. The simulations provide insight into the separation mechanism and allow the optimization of specific devices depending on the application. Good agreement is observed  with microfluidic experiments measuring the deflection of polystyrene particles, white and red blood cells, and MCF-7 breast cancer cells. Promising applications of this platform include the fractionation of blood and the capture of circulating tumor cells. This platform also has the potential to yield  samples with controlled polydispersity at high throughputs, which would allow the study of the rheological behavior of complex fluids with  desired particle size distributions.