PDMS+Shrinky Dink Valves

Here I explored the possibility of sandwiching an elastomeric layer between polystyrene layers to make valves.

I started by casting thin layers of PDMS to use as the in between layer. I used Sylgard 184 and aimed for a slightly higher mix ration (7:1 in stead of the usual 10:1) in the hopes of making a slightly more flexible membrane. Four membranes were cast into petri dishes with various thicknesses below 1 mm.


Quick Notes of PDMS:

  • Mix PDMS really, really well, especially with such thin layers. You should degas it before poring but don't worry about little bubbles on the suface too much. They will pop by the time it's done curing. If you have bubbles stuck to channel features, either degas the whole thing or use a hollow needle to pop them carefully.
  • Sylgard 184 v RTV-615 v others: I tend to prefer Sylgard 184 when working with things like cells and such. It is in theory "cleaner". RTV-615 is somewhat easier to plasma bond though and may be better for multilayer devices. There seem to be a bunch of other brands that claim to be the same as Sylgard 184 at least for solar cell encapsulation. In some cases these alternatives appear to be significantly cheaper. Might be interesting to test these.

  • A simple test platform was drawn up as seen below. The vertical channel was used as the control channel and the horizontal channel as the flow channel. Layers of PDMS were then sandwiched between those two layers.


    The thinnest PDMS membrane (about 0.25mm) was most successful at closing the flow channel but also ruptured after a couple of runs (before pictures could be taken). One possible solution going forward is to make the channels dimensions wider and shallower relative to each other so the PDMS has to deform less in order to close the channel. Also interesting would be characterizing the channel geometries as flow channels in these valves are usually rounded while the control channels are rectangular.