Since the substances should be distributed over a large area and evenly on the skin, I designed a microfluid pad. For this purpose, a corresponding structure was built using ABS filament and then cast in 2-component silicone. Next, it was degassed by a vacuum chamber. After curing, the cast was placed in acetone. The acetone dissolves the ABS filament but does not attack the silicone.

After one day in the acetone solution, the channels were flushed using a small syringe to remove the dissolved acrylonitrile butadiene styrene. Then the silicone mold was placed in fresh acetone solution for another day, followed by a few hours in a 10% hydrogen peroxide solution. Finally, the silicone mold was rinsed several times using hot distilled water.

In addition to the microfluid pad, the system consists of a 6V peristaltic pump and a reservoir – a modified liquid bag for airplanes with a capacity of 30ml.

A three-way valve controlled by a servo would also be desirable. Then you could use two reservoirs and even mix substances. There are small and cheap medical three-way valves that are actually intended for infusions. It should not be too difficult to hack one of these.

Since the plug of the three-way valve is pressed in, it is somewhat stiff. The ES08 servo has too little torque for this. Therefore I took a Hitec HS-85 MG servo for the prototype. I used 3x2mm stainless steel tubes for the three ports, which were glued with 2-component epoxy glue. The two mounts were cut out of a 2mm GRP plate. These GRP plates are excellent for prototyping, as they are very easy to work with and extremely sturdy.

The microfluid pad cannot be placed directly on the skin, as the skin would close the channels. After experimenting with different materials, a cellulose sponge cloth turned out to be the most suitable. Using an appropriate hole punch, I punched out a disc whose diameter corresponds to the microfluid pad. This works best when the sponge cloth is damp.