In Idea #119, we proposed to filter atoms used slabs of a material that expand very little with temperature. The idea here has the same goal, to produce an atomic filter, but uses a different approach. 

Imagine that you have a flat paper surface. Then you take an ultra-sharp pencil, or rather a graphite rod, and start drawing a trajectory into the paper, without lifting the carbon rod. A random trajectory, it can be a zig-zag pattern bouncing off the edges of the paper or circles with various radii, or just a curly random line. 

The paper starts to get filled with carbon. The white areas start to slowly disappear, and become smaller and smaller.

If you keep this process for a long time, do you end up with a filter of atomic sizes?

Is this filter dense in holes enough to be useful?

By the way, you are supposed to dissolve the paper after it has been draw to. 

You can also try this by pouring on the surface a trickle of something that solidifies in contact with a surface, like a polymer, and moving the container you are pouring. I suppose you could also use droplets, anything to fill most of the surface.