Creating Droplets in Microfluidic Devices with Ultraviolet Light

Academic Research
Digital Microfluidics
Damien Baigl
RSC Lab on a Chip
École Normale Supérieure



August 6, 2011

Digital Microfluidics Background

With the widespread use of electronics, we often use the word ‘digital’, but we might not always think about what it actually means. For those of you who have never taken a class in electrical engineering, or never learned Latin (from the word digitus), the word describes anything that is discrete as opposed to continuous. Digital has also been applied to a type of microfluidics. With the definition in hand, you might guess that digital microfluidics does not describe continuous fluid flow through channels at the micro- scale, but instead is made of droplets. Discrete droplets can be implemented in a variety of assays or devices, as they would allow for complete control of the fluid, instead of a continuous stream of fluid that may not be thoroughly mixed. Regardless of the intended use of the droplets, they must first be created.

There are currently three techniques to generate droplets:

  1. electrowetting
  2. dielectrophoresis
  3. emulsification

Electrowetting essentially uses an electric field to change how fluid interacts with the surface. It can make the surface more or less attracted to water, causing a fluid such as water to ‘hug’ the substrate or ‘ball up’ into a droplet. Manipulation of the electric field would provide control of the locations of droplets and how they move. Check out the videos from Dr. Richard Fair’s laboratory at Duke that illustrate the formation and transportation of droplets using electrowetting.

Dielectrophoresis occurs when nonuniform electric fields cause polarizable particles to move. The application of dielectrophoresis for microfluidics was proposed by Dr. Thomas B Jones in the Journal of Applied Physics in 2001. Water is attracted to the regions where the electric field is the strongest. This movie from Jones does not demonstrate the formation of droplet formation, but it does illustrate its control over water.

Finally, the process of emulsification describes a system of two fluids in which one fluid is dispersed throughout the other. Think water and oil and the droplets you can create when you shake it around. This first requires at least two fluids to be used (I say at least two, because multiple emulsions can be achieved, as seen here) and an external stimulus. An external stimulus is often needed to cause stable droplet formation. This can occur at junctions, where specific geometry, along with control of flow rates can cause emulsification. You can see a video at the company RainDance’s website. They have more information on the subject, and I recommend that you check out their other videos on that page, especially ‘Loading droplets’. It’s like a gumball machine!

UV Controlled Droplet Formation

While the ability to digitize fluid is valuable, its regulation can be increasingly more prized. Does the digitization have an on/off switch? Do you have to change the flow rates of the system to revert back to continuous flow, or physically move components that are responsible for pinching the droplets? The capability to switch between continuous and digital flow could serve to reduce the footprint of the device. Why make the system larger just to incorporate streams and droplets when it can happen in the same place? This would lend elegance to the design of the device. The sophistication would be improved if this could be accomplished without moving parts. The most reliable instruments and devices have fewer moving parts that could break down. This might be what Damien Baigl et al. from École Normale Supérieure in France had in mind. Their research, which was featured on the cover of the 2011 Issue 16 of Lab on a Chip, proposes a method to emulsify droplets with Ultraviolet (UV) light. Their paper entitled, “Photoreversible fragmentation of a liquid interface for micro-droplet generation by light actuation” describes an emulsification system that is controlled by the use of UV light.

To start, the system has water-in-oil flow. But the water contains a surfactant AzoTAB. When UV light is applied to this compound, a double covalent bond switches (from trans to cis). This causes the surfactant to become more polar and decrease the wettability with the surface of the device. The lowered wettability causes the water with AzoTAB to form droplets. Initially the researchers created a junction that was capable of emulsification depending on the flow rates of the fluid. They were also able to find a combination of flow rates that would not normally create droplets, but digitized with UV light.

While this is a nice feature, it partially relies on the preexisting structure that is capable of emulsification. They next presented a design that does not constrict the fluids. This was also able to cause droplet formation with UV light. It was demonstrated that the presence and absence of UV light resulted in digital and continuous flow. This partially fulfills the desired versatility I discussed earlier. But I think that in addition to being able to generate droplets and streams at whim, it is also advantageous to convert droplets back into a stream. The authors briefly mention this, and it seems that the application of blue light can reverse the switch and produce streams. I think the greatest part of this setup is the ability to change the same unit of fluid between continuous and digital.

But what does this research really get us? Well, nothing at first. This isn’t a complete device like Dr. Sam Sia’s mchip that can detect HIV. But this will surely be incorporated into a device. It really is a tool that can be applied in different ways along with other tools to create a full device. I’ll update you further once this has been incorporated and used further.


Diguet, A., Li, H., Queyriaux, N., Chen, Y., & Baigl, D. (2011). Photoreversible fragmentation of a liquid interface for micro-droplet generation by light actuation Lab on a Chip, 11 (16), 2666-2669 DOI: 10.1039/C1LC20328B

Videos reproduced by permission of Damien Baigl and The Royal Society of Chemistry from Lab Chip, 2011, 11, 2666-2669, DOI: 10.1039/C1LC20328B.