Overview:

In Summer 2022, I interned at the Keck Graduate Institut'e Aran Lab which specializes in developing diagnostic biosensor chips. These chips can detect trace amounts of any protein or viral particle. My task was to design low-cost microfluidic flow cells that could facilitate mass-testing of chips, create a QA database website using Bootstrap and SQL, and run CAD + CFD workshops for the Undergraduate Medical Device Bootcamp. We pitched these microfluidic devices to Cardea Bio Inc., the producer of these chips, and they expressed interest in mass-producing our attachments for future use in testing their biosensors.

Project 1 - Microfluidic Flow Cells:

The purpose of these microfluidic devices were that they had to:


1. Have an inlet and an outlet for solutions
2. Locate themselves to align themselves in large assays
3. Allow for a film to be placed over the sensing reservoir and still let liquid through


Our initial design utilized angled pipette inlets and air outlets, which later proved to be not user-friendly. We also used interlocking dovetail systems to keep multiple attachments and sensors togther while testing large assays. However, this caused accuracy issues. The buffer solution also got stuck easily on the unfilletted edges of the top chamber.

In our new design, we implemented vertical inlets with angeled pathways instead of angled inlets. This greatly improved the continuous flow of liquid. We also put radiuses on the top chamber and got rid of the dovetail features. Instead, each attachment was datumed to the bottom edge of each chip.

Project 2: QA Database Website

My second project was to create a website that allowed for easy input and access of QA data for wafer testing. Each wafer has over 70 sensors. I designed a graphic that represented the wafers being tested and made it interactive where the user performing the QA could input a name and data for each of the sensors by clicking on its representative square. This data would then be stored in a library of wafers that the researchers could easily access later on. I also developed a log in page to make it more secure. This website was built using Bootstrap for the UI, SQL for the database, and Python to make everything interactive and cohesive.