Formation of monodisperse bubbles in a microfluidic flow-focusing device

From Soft-Matter
Jump to: navigation, search

Original entry: Warren Lloyd Ung, APPHY 225, Fall 2009

"Formation of monodisperse bubbles in a microfluidic flow-focusing device"
Piotr Garstecki, Irina Gitlin, Willow DiLuzio, Eugenia Kumacheva, George M. Whitesides and Howard A. Stone.
Applied Physics Letters (2004).


Soft Matter Keywords

monodispersity, microfluidics, bubbles, multi-phase flow

Figure 1: The microfluidic device for generating bubbles
Figure 2: Lattice flows of bubbles of various sizes
Figure 3: Graph of bubble volume v. the product of flow rate and viscosity
Figure 4: (a) Graph of the frequency of breakup v. liquid flow rate, (b) radius of bubbles formed (circle), volume fraction of bubbles in channel, <math>\phi_{ch}</math> (triangle), and volume fraction of bubbles at formation, <math>\phi_{or}</math> (square).

Summary

A microfluidic device was created in which gas bubbles are generated within a liquid flow using a flow-focusing technique. The device works by creating a flow of gas within a continuous liquid phase and forcing it through a narrow orifice. The gas is focused into a thread which passes through an orifice, but there is instability produced due to the gas thread's large amount of surface energy. The gas thread breaks off and the gas in the output channel collapses into a bubble to minimize its surface energy (refer to Figure 1a). Bubbles can be formed at rates exceeding 100kHz. The geometry of the device is such that bubbles are constrained by both the top and the bottom of the channel, but are unconstrained along the width of the channel (shown in Figure 1b on the right).

Operation of the device within a certain regime produced a single, monodisperse bubble for each break in the gas thread. The device was also able to operate in a regime in which two bubbles were formed before a break was observed. The two bubbles are of different sizes with respect to each other, but the set of two bubbles was monodisperse with respect to the other sets of two bubbles formed.

Soft Matter Discussion

By varying the pressure of the gas, the flow rate of the liquid and the viscosity of the fluid, it is possible to control the volume of the bubbles formed as well as their frequency (see Figure 4a). The overall relationship between volume, pressure, flow rate, and viscosity is as follows:

<math>V \propto \frac{p}{q\mu}</math>

One significant observation was that the surface tension of the liquid had no bearing on the volume of the bubbles generated; it only affected their tendency for coalescence following generation. This was tested by removing the surfactant from the liquid phase, in effect doubling the surface tension. Bubbles without surfactant merged upon contact to form large bubbles, whereas bubbles with surfactant retained their form and were able to flow in complex lattice arrangements of bubbles (see Figure 2). It is postulated that these lattice arrangements exist in these configurations because the top of the channel must deform outward due to the pressure of the flow creating a rounded surface to which the bubbles can conform, lowering their interfacial area.

Another interesting observation is that if volume, <math>V_b</math>, is plotted against the product of the flow rate, <math>q</math>, and the viscosity, <math>\mu</math>, then the volume curves for all variations of flow rate, viscosity, and surface tension follow the same trend (see Figure 3). Note that this variation takes place while holding the gas pressure constant and in a set of microfluidic devices with fixed flow parameters (<math>w_{or} = 75\mu m</math>, <math>w_{out} = 750 \mu m</math>, and <math>L = 30 mm</math>).

It is also shown that the volume fraction at which the bubbles are formed, <math>\phi_{or}</math>, is not the same as the overall volume fraction of bubbles in the output channel, <math>\phi_{ch}</math>. This is explained as arising from a difference in speed between the bubbles and the liquid phase due to viscous dissipation at the thin film between the bubbles and the walls of the channel. This can be confirmed since the volume of the bubble, the volume fraction of bubbles in the channel, and the frequency at which bubbles are formed are all known.