A membrane with small pores and air on the other side could do the trick because you will have to overcome a capillary pressure. However, this is only going to work once. After that there will most likely be a wetting layer at the outside of the membrane which effectively reduces the capillary pressure drop over the membrane to 0. To make this work you need to dry the outside of the membrane every time, which is very impractical.
What I would suggest is to use something similar to an elastically deforming membrane like the one you find in beakers for small children You could use a small buffer region in the valve to make sure that the flow is somewhat even. Entirely even is not going to be possible, because you will most likely have flow rates that are low enough to have laminar flow such that your profile will be parabolic