Research

Microfluidics deal with the behavior and manipulation of fluids in channels with dimensions of sub-millimeter scale. Sorting and separation of particles are key elements in many microfluidic processes. Previous studies focusing on spherical particles and droplets may not be applicable to the manipulation of compliant biological cells and synthetic capsules in microfluidic devices. First, we developed a numerical model combined the lattice Boltzmann method (LBM) for fluids and the mass-spring model for solids to study the dynamics of deformable capsules in microchannels. Inspired by this fundamental study, my collaboration with the Sulchek BioMEMS Lab turned out a new design of a microfluidic device for continuous, label-free cancer cell separation and detection by mechanical stiffness. We investigated the underlying mechanism of cell separation through numerical simulations, which were in good agreement with experimental results. This sorting method would greatly complement existing separation techniques, which are critical for detecting diseases such as cancers and sickle-cell anemia.

We use computational modeling to design a hydrodynamic method for high-throughput separation of solid microparticles by size in microchannels. The rapid and high-resolution separation occurs due to a combination of two hydrodynamic effects: cross-stream inertial migration of particles and circulatory fluid flows created by periodic diagonal ridges protruding from opposite channel walls. This new continuous separation method operates in a wide range of Reynolds numbers, is insensitive to the magnitude of channel flow rate, and features simple design that can be readily integrated into microfluidic devices for massive sample analysis.

In nature, biological and synthetic particles are normally non-spherical. Therefore, I extended my work to the dynamics of spheroid particles in a flow. Although the rotation of spheroids in shear flows has been widely studied, a comprehensive understanding of the combined effect of fluid and particle inertia on the orientation of spheroids with different aspect ratios is yet to be established. Our numerical study filled the gap and systematically investigated the complex dynamics of spheroids in a shear flow, which deviates from Jeffery’s orbits due to the inertial effect. The results and techniques used in this study can be applied for designing microfluidic devices for separation of non-spherical particles. Indeed, collaborated with the Dr. Di Carlo's Lab at UCLA, we explored that this high-throughput shape-based separation can be applied to a large range of particle sized and types, including artificially made 3-µm particles as well as bioparticles such as yeast

Micromixing is a key technology to fields such as pharmaceutical industry, analytical chemistry, biochemical analysis, and high-throughput synthesis since it reduces the quantity of agents involved in the chemical and/or biochemical processes. My research in this field expanded the understanding of rapid mixing and facilitated the design of two active micromixers. One of the design is to utilize pulsating square waves introduced at the entrances of a simple T-type mixer. This process significantly increases the interface area of fluids allowing them to reach a uniform mixing rapidly. Key parameters, such as the Reynolds number, the Strouhal number, and the amplitude of pulsatile wave and their compounding effect on the mixing efficiency were thoroughly investigated. Collaborated with Dr. Peter Hesketh at Georgia Tech, we devised a novel mixing method by the manipulation of magnetic beads. Previous electromagnetic micromixers are limited to the manipulation of a small number of magnetic beads due to Joule heating. Our device overcame this limitation by using soft magnetic patterns on the microchannel wall to control the motion of a large amount of super-paramagnetic beads. Through CFD and magnetic field modeling, I deduced the best mixing parameters for achieving the most effective mixing.

Thermocapillary flow or Marangoni convection is the mass transfer along an interface between two fluids due to surface tension gradient induced by the temperature difference. For open microfluidic devices, free-surface flows can be actuated by thermocapillary convection. We performed CFD simulations (the volume-of-fluid method) and the long-wave theory (collaborated with Dr. Alexander Oron at Technion) to study the thermocapillary flow arising in a thin liquid film on a non-uniform heated solid substrate. An efficient fluid pumping mechanism was discovered. In a similar scenario, when the Marangoni number exceeds the threshold value, the film breaks into a series of drops with minuscule fluid bridges connecting each of them. By increasing the wave frequency, the bridges thicken and allow fluid to transfer between the drops. Both of these studies revealed new free-surface flow techniques in microfluidics and provided engineers with more tools in the design.

While it is known that eyelashes serve as protection and prevent eye infections, the correct physical explanation has been unknown. Collaborated with the Dr. Hu’s Biolocomotion Lab at Georgia Tech, we performed a study on the aerodynamic characteristic of eyelashes using physical experiments in a wind tunnel and numerical simulations. We found that eyelashes diverted airflows, thereby acting as a passive dust controlling system for eyes. Most mammal eyelashes are the optimum length since any longer eyelashes would funnel particles into eyes. In a separate study, I applied similar techniques to the ocular hairs found on insects. Just like mammal eyelashes, insect ocular hairs were proven to have an optimal ratio, with hair length equaling their spacing. This study revealed that an array of hairs reduces surface airflow by 90%, effectively creating a shield of stagnant air for a flying insect. These studies not only elucidated the long mysterious function of eyelashes, but also shed light on a new perspective of the dust cleaning system.