Got data for opacity
there, hope it's correct, telling that:
- Aluminium is better
than gold in the visible spectrum
- 20nm of Al leave 7% at 800nm, 5% at 400nm and 6% at 171nm. Hole definition is efficient, background light is too strong, you're right.
BUT we can have a thicker layer everywhere and leave it thinner locally to keep the contrast. For instance, if the hole is made by pressing the tip through the soft metal
against the hard ceramic, then a conical hole
can be narrow at the bottom (let's keep D=20nm) and wider at the top. At 90° full angle, only D=60nm have <20nm thickness, and D=100nm have already 40nm thickness, and so on. This looks simple and acceptably efficient.
If the hole is punched by pressing the tip through very pure aluminium (which is ultra-soft), the harder ceramic stops the movement, and this defines a narrow hole. The other previously suggested processes must work as well.
Immediately around the hole, aluminium isn't very opaque, but enough to define a diffraction pattern; farther away, it's thick enough to avoid background light.
Marc Schaefer, aka Enthalpy