(248i) Enhancing Bioparticle Trapping at a Converging Micro-Flow with Local Coulombic Forces and Roughness-Induced Surface Currents
AIChE Annual Meeting
Tuesday, November 6, 2007 - 2:30pm to 2:45pm
We had recently reported a new microfluidic technique that can rapidly concentrate a dilute bio-particle (bacteria, blood cells, immuno-beads) suspension with particle count as low as ~ 104/ml (D. Hou, S. Maheshwari, and H.-C. Chang, "Rapid bioparticle concentration and detection by combining a discharge driven vortex with surface enhanced Raman scattering," Biomicrofluidics, 2007). This concentration technique relies on a long-range converging spiral flow to rapidly concentrate particles into a packed mound on the bottom substrate of a liquid chamber. The spiral flow is produced as a combination of a primary azimuthal vortex and a secondary radial flow due to inertia. The resultant flow generates a converging stagnation point at the bottom of the chamber which serves as a stable focal point for concentration. The particles follow the fluid streamlines till near the stagnation point, where the velocity becomes very low as compared to the bulk flow. If viscous drag can be overcome, the particles can be dislodged from the streamlines by gravity or additional body forces to accumulate at the stagnation point. We estimate this condition with a dimensionless parameter l that balances the trapping gravitational forces with the re-suspending viscous drag on a single particle. Our numerical simulations confirm this hypothesis, and we find a narrow window for optimum trapping when l is of unit order. When l is small, viscous drag re-suspends all the particles near the stagnation point and bulk chaotic mixing after several circuits around the spiral flow disperses them uniformly within the vortex. This mixing effect results in particle deposition over a large area around the stagnation point. This is particularly problematic for bacteria and other bio-particles with very low settling velocities due to small dimensions and densities comparable to the medium. Uniform sedimentation also occurs when l is large such that radial convection of the particles is weak and an initially well-mixed suspension produces uniform deposition.
To enlarge the trapping window for small l values, an additional downward-acting coulombic force at the stagnation point, e.g. dielectrophoresis(DEP) or electrophoresis (EP) is imparted, with micro-electrodes fabricated near the stagnation point. Augmentation of the gravitational force with additional coulombic force shifts the trapping from the small l region to the unit l region. For the large l situation, we generate a surface current by roughening the bottom substrate by fabricating well-regimented micron-sized metal or glass pixels of height h and separation w at the bottom of the substrate. Roughness prevents the settling of the particles by producing an additional secondary local flow around the pixels, the magnitude of whose normal component is proportional to the product of the primary shear rate with (h2/w). Balancing this roughness velocity with the particle sedimentation velocity allows us to obtain a lower bound on (h2/w). Adhesion and steric considerations then stipulate that both h and w should be comparable to the particle size. There are hence two trapping regimes; one corresponding to particles being trapped directly from the bulk at short times and another corresponding to the surface roughness flux at longer times. The geometry of the micro-pixel array is found to have little influence on the surface flux. However, specific pixel dimension and separation allows us to concentrate particles of particular size and density, thus converting the stagnation flow into a particle segregation device. We have experimentally confirmed the extension of the trapping window towards both small and large l values, consistent with the results obtained from a scaling theory. The effect of the applied voltage and frequency on trapping of particles of different dimensions and the influence of the aspect ratio of the trapping chamber has also been analyzed. Quantification of these observations has provided the necessary design criteria for optimum trapping under different conditions.