Development of an Efficient Quasi-3D Microfluidic Flow Model and Fabrication and Characterization of an All-PDMS Opto-Microfluidic Flow Cytometer.

摘要

In this thesis, development of a novel microfluidic flow model, and, fabrication and testing of microfluidic cytometer for potential cell detection and sorting applications are described. The model is formulated by decomposing the flow profile along the height of microfluidic device into a Fourier series that converts the 3D flow equations into a series of coupled 2D equations and is applicable to planar microfluidic devices only. It is validated against the analytical solution for flow in a straight rectangular channel and the full 3D solution of a commercial Navier-Stokes solver for flow in a T-channel. Comparable accuracy to the full 3D numerical solution is achieved by using only three Fourier terms with significant decrease in computation time. The model is also extended to the problems with time-varying boundary conditions.;We fabricated two first generation miniaturized cytometer prototypes and used them for preliminary proof-of-concepts experiments. They were built by cutting fluidic channels into two different polymer materials and bonding them between two standard glass slides with epoxy and fusion bonding.;We fabricated a second generation of flow cytometer chip consisting of an integrated 2D hydrodynamic focusing system, solid-core optical waveguides and a hydrodynamic side-flow switching system on an all-PDMS platform.;Optical propagation losses of the integrated waveguides and signal-to-noise ratio (SNR) of its detection system were characterized. The propagation losses were found to be 1.6 and 1.5 dB/cm for the green and red light, respectively. Detection of fluorescent signal through the waveguide yielded improved SNR than the conventional method of under-chip detection.;Fluid flow speeds were estimated from volumetric flow measurements and fluorescent particle tracking experiments and the width of the hydrodynamically focused stream was extracted from microscope flow images. The results were compared to the simulation values obtained from the Q3D model and reasonable agreement was observed. Detection and sorting of microparticles were demonstrated using this device and initial results are presented.;The numerical model, the fabrication techniques, and the experimental methods developed in this thesis may be applied to many biomedical engineering applications that use devices utilizing microfluidic flow and optical interrogation.