Entrapment of bubbles in microfluidic devices can distort the fluid flow and negatively impact device performance. For example, bubbles can obstruct the formation of jets in inkjet printers and can cause faulty reporting for sensors embedded in microchannels. In some cases, the presence of bubbles cannot be completely avoided. In these cases the bubbles ideally should flow through the channel without entrapment. The likelihood of bubble entrapment depends on the geometry of the channel, the flow characteristics, and the surface properties of the microchannel walls.
To solve this problem, we developed a multiphase CFD model (using COMSOL Multiphysics) of a bubble moving through a 1 mm x 2 mm channel featuring a mild restriction in the middle, and a bubble of 0.3 mm radius initially located towards the bottom of the channel. We accounted for surface tension between the bubble and the surrounding fluid, and the surface properties of the channel walls (hydrophilic, neutral, or hydrophobic). We assumed the channel walls to have a neutral (90º) contact angle with the bubble except at the constriction.
The animations in Figure 3 show the travel of the bubble for a contact angle of 22.5º between the bubble and the channel walls in the constriction. The bubble contacts the channel wall, and moves upwards until it gets trapped close to the end of the constriction. The fluid is forced around the bubble, creating a non-symmetric velocity field.
The set of animations in Figure 4 show the same configuration, but with a 90º contact angle throughout the channel. The bubble in this case does not get trapped and eventually exits the channel. Note that this model does not account for contact angle hysteresis (different advancing and receding contact angles) which increases the likelihood of bubble entrapment.
The average speed of the bubble over time for both configurations is shown in the graph in Figure 5. The increase in speed between 0.05 and 0.1 seconds is due to the reduction in channel area as the bubble passes through the constriction.
This example involved a simple microchannel geometry to illustrate the concept. We used this approach for a more realistic geometry where the simulation provided design insight on the range of geometries and contact angles that can cause bubble entrapment.