My advice on this would be to add a very low flow of Oxygen. I have quite a
bit of expereince plaama etching oxide films, and very thick films are the
worst due to the polymer formation,redeposit situation. Typically a pure cf4
flow should not be a polymern issue, so this is a bit perplexing. The oxygen
consumes carbon in the reaction, keeping the critical carbon-fluorine ratio
outside the polymer formation point. The general rule of thumb for these
higher pressure RIE (parallel plate) type processes is that if the c/f ratio
is 1/3 you are fine. If it goes to 1/2 you are on the edge. If the issue is
selectivity to an underlying Silicon layer you have to walk that tightrope,
but diampnd should have better selectivity-this I am guessing. The diamond is
a carbon film, but I have no idea whether the crystal structure would keep the
carbon constrained. Adding oxygen is not good for stopping on carbon,
obvioulsy. Alos, keep the flow very low, in your mix maybe 3 sccm or less. The
oxygen can be a sputtering enhancer, which would attack the aluminum, and
result in aluminum-oxide which will really deposit. Some competting reactions
and mechanisms here. With high pressure RIE, you are better off playing with
chemistries than trying to control the physical aspect of the etch.
Good luck