Sean Harbertson, University of Utah
The purpose of my research is aimed at establishing the ideal conditions for micro scale particle filtration on a microfluidic spiral channel device. Sperm washing, or filtration of sperm cells from semen, is a very important step in the process of artificial insemination. Before a sperm sample can be artificially inseminated it must be filtered of the seminal plasma present in the sample of semen. Current methods usually involve filtration by density using a centrifuge. This method, although proven effective, is damaging to sperm cells. Our goal it to improve upon this method by utilizing microfluidic devices with spiral channels as a method of filtration that is faster and isn’t as damaging to cells. This filtration technique works due to curvature of the channels in combination with the principals of inertial lift and Dean drag forces. These conditions result in partials of different diameters lining up at different positions along the channels width allowing filtration of partials due to size. To model this system we use a mixture of deionized water and micro-fluorescent beads that are 3 microns in size to represent the seminal plasma and sperm cells. The bulk of my research has been to collect data images of the separation observed when mixtures of deionized water and the fluorescent beads are run through the device at different speeds and concentrations in order to determine the speed and outlet size that results in the fastest filtration time. The next step was to evaluate each of the images collected to determine how wide the collection channel needs to be in order to collect the cells while minimizing the amount of fluid collected with the cells. Using the recorded speed at which the fluid was pushed through the device and the data collected from processing the images, the last step will then be to calculate the ideal speed to push fluid though the device to maximize filtration while minimizing the time it takes to completely filter a sample. From the data collected and evaluated thus far, we expect that ideal speed at which to run a sample through the device will be around .5 ml/min. We also expect that with adjustments to the collection outlets of the device the total time to filter an average size sample will take less than 20 minutes.