Droplet behaviour in microfluidic devices

  • Pinar England

Student thesis: Doctoral Thesis

Abstract

This work is a study to understand the various aspects of a microfluidic device. In the first half we take the role of an end user, experimenting to learn how best to use the device efficiently. In the second half we are the manufacturer, trying to fabricate a user friendly, and fully functioning microfluidic device. As the end user, we have three different T-junction droplet generator devices, with similar geometries. We start investigating by generating water droplets in an oil medium. They self-organise into various flow patterns: single-profile, double-helix profile and triple-helix profile. We document how, with increasing flow rate ratio and capillary number, we observe more densely packed droplet flow patterns. The device with the deeper expansion channel provides more space for the droplets and they self-organise the triple-helix pattern in 3-dimension. We then use the same devices to generate droplets for which we can calculate the volume. The fluid flow in a microchannel happens in four different regimes: ballooning, squeezing, dripping and jetting regimes. In single-cell and single-molecule analysis devices, the ability to create droplets on demand and of a certain volume is a desired capability. This can be achieved by understanding and learning how to use the fluid flow characteristics accurately. We experiment with the three different sized microfluidic devices, to measure the droplet volume throughout the squeezing to dripping regimes. This is achieved by manipulating the capillary number and the flow rate ratio. We observe a similar result as with the flow patterns: that the capillary number has an impact on the droplet volume. As the capillary number increases the droplet diameter decreases. Further, for a set capillary number we can fine tune the droplet diameter by changing the flow rate ratio. As the flow rate ratio increases the volume of water droplets increases, despite the fact the capillary number is set.These coincide with our flow pattern results. Our results fit to the scaling law to predict the droplet size introduced by Tanet al. in 2008 [51]. Unlike some other authors in the literature, we did not observe a critical capillary number where the droplet volume changes suddenly. However, we did observe a transition area where we cannot define the regime of the fluid flow. As the manufacturer we designed and fabricated our own planar free standing microfluidic devices using a polymer called SU-8. After looking into the weaknesses and the strengths of using SU-8, we describe how we successfully fabricated working devices and developeda new procedure in adhesive low temperature bonding. We finish by considering the challenges of connecting micro sized structures to a macro sized syringe pump, and fabricated a chip-holder inspired by applications in industry.
Date of Award1 Jun 2011
LanguageEnglish
Awarding Institution
  • University Of Strathclyde
SupervisorYonghao Zhang (Supervisor) & Matthew Stickland (Supervisor)

Cite this