Some applications in (aqueous in oil) droplet microfluidics require the encapsulation of viscous liquids, e.g. high concentrations of reagents, polymers, or gelling agents such as agarose, alginate or gelatine for gell microbeads. Several publications mention a higher polydispersity (a larger range of droplet sizes) when generating droplets from viscous liquids, and I have seen many different droplet generation junction designs for PDMS chips being used at different flow rates (e.g. 2D T-junctions, 2D flow focusing with various channel geometries of aqueous channels and oil channel angle, 3D designs with a step). Which junction design is most suitable to generate highly monodisperse picolitre droplets of viscous liquids at the highest suitable flow-rates (high throughput), and why?
Thank you Adam Opalski , I see that often there is no best design for all applications. I think what I am looking for, however, is a solution to a widespread problem: I am looking for a PDMS based design (most common and most flexible manufacturing method, not e.g. co-flow in capillaries) that is relatively easy to make (2-layer chip is OK, but aligning several PDMS moulds as well is a fabrication nightmare), and that has a good quality (here specifically for 0.5% - 2% gels) as well as throughput (some papers report homogeneous droplets but generate them at a throughput more than two orders of magnitude lower than low viscosity droplets - not practical).
I know the paper you mention, but I am not convinced by it's practicality. Our group did not yet obtain monodisperse droplets from the last, similar parallelised design published by the senior author, and there is little rationale in the paper why their design would work well (the hydrodynamic pushing does not make intuitive sense to me as a physicist - it should be suction instead? + very little flow visible in their videos). More importantly, their design is very difficult to manufacture (two layers plus alignment of two chips).
There is however this paper, that gives some rational as to why their junction design with flow focusing puls step works well:
Article Monodisperse alginate microgel formation in a three-dimensio...
I see that you have a background in step-designs as well. What is your opinion on this method?
Hi Tobias,
Take a look at the soft lithography methods that the Kumacheva and Simmons Labs at U of Toronto uses. Article Simultaneous generation of droplets with different dimension...
This may help to guide your decisions.
They did studies involving making gels and other viscous monodisperse droplets.
Good luck
mike
I agree that the step emulsification approaches are more simple to scale up due to their scale invariance. The 3d designs in the first paper mentioned are probably not necessary, a simple 2 layer design like previous work published with step emulsifiers would suffice.
We recently published some work using a step emulsifiers to fabricate hydrogel particles in high throughput (https://onlinelibrary.wiley.com/doi/10.1002/adfm.201900071). This was with 12% 20kda peg precursors. We have also used these with gelatin methacrylate heals up to 20wt% with great success. In both cases we achieve orders of magnitude higher throughputs then flow focusing geometries. With the more complex and viscous materials flow focusing will always be difficult to scale. Although for more simple fluids researchers have successfully demonstrated high parallelization with high uniformity ( e.g. Issadore nature comm. 2018).
Hi Tobias,
First of all we should know how much CV% (coefficient of variation) you need when you are talking about droplet monodispersity. For example if you need to have droplets with 1% CV then it is required to work with high quality flow sources. In general, droplet monodispersity is highly influenced with pressure or flow rate fluctuations that induced to the chip. It does not matter with which microfluidic device or fluids you are working, when there is high fluctuations in the pressure or flow rate you will end up with polydisperse droplets. Therefore, the important thing is the flow source. If you want to change flow rate for droplet generation you need to use syringe pump that normally induces high fluctuations of flow rate to the system (due to the mechanism of the motor step inside syringe pump). Then, to solve the problem you can utilize compliance effect of soft tubing or soft PDMS to decrease those fluctuations (Article Damping hydrodynamic fluctuations in microfluidic systems
) (Article A soft microchannel decreases polydispersity of droplet generation
). I could get CV=0.48%, using long flexible tubing. For the pressure pump, you need to have proper design of hydrodynamic resistance and work in high pressure (Article Global network design for robust operation of microfluidic d...
). After all, I should say you can make highly monodisperse droplets with a simple T-junction device if you can decrease pressure or flow rate fluctuations as much as possible. Good luck.
Hi Tobias,
It is not easy to generate monodisperse droplets from viscous inner phase (for example, alginate or agarose) with droplet microfluidics, since the droplet generation process is usually jetting and finally develops into sheath flow. However, I have read a research paper where phase inversion was applied to generate such droplets. First, the viscous phase was used as the outer phase, and droplets of a non-viscous phase was generated. Then, the hydrophilicity/hydrophobicity of the microfluidic channel changed, so that the plug-shaped droplets underwent phase inversion, and droplets of the viscous phase were formed. Hopefully this helps.
- Badges
- Science topic
- Similar topics
- Engineering
- Modeling