[mems-talk] Help with RIE chemistries and removing hard baked Photoresist.

[Martyn Gadsdon]

Dear All,

What I have been recently attempting to do is to spin a layer of organic 
polymer (something that will be etched by an O2 RIE plasma) onto a 
substrate, then on top of this evaporate a thin barrier layer, and then spin 
photoresist on top of these two layers. The idea being that once I have 
created a pattern in the upper photoresist (feature size ~ 200nm) I can etch 
this through the barrier layer exposing the base layer below. Therefore, 
after etching the barrier layer with one chemistry, I can use a second RIE 
of pure O2 to etch through the base layer exposing the substrate with the 
desired pattern. Therefore I need a barrier layer that can be etched either 
without using any O2 in the plasma or at least, if it needs to be present to 
prevent grassing as with CHF3, will etch at a comparable rate to the 
photoresist on top of it; furthermore, it must not be etched using a pure O2 
plasma. I then evaporate a metal on to the remaining structure and 
subsequently wash away any remaining organic layer and hence barrier layer 
and unwanted metal, hopefully leaving only the desired pattern of metal 
behind on the substrate.

So far I have found 20nm Silicon to be a good barrier layer because it will 
etch at approx the same rate as photoresist in a CHF3/O2 plasma if the 
relative gas concentrations are correct, and it will not be etched by pure 
O2. However, for my base layer I have been using hard baked photoresist 
because the Silicon needs to be evaporated at 1410 Celsius. Unfortunately, I 
am having a really hard time trying to remove the underlying photoresist.

So my first question is will the recipe for removing PR that has been in a 
RIE, which was recently posted, also work for really hard baked PR and will 
it affect the metal? And if not, are there any other wet etch recipes that 
will work but not affect metal?

The recipe:

"place the sample a fwe min into hot acetone (50C), then few hours hot
N-Methyl-2-Pyrrolidon (NMP). This is the basis of most PR. Then hot acetone
again and then Isopropanol. "

Secondly, I have tried using PMMA as a base layer because this is much 
easier to remove. However, I have found that when the silicon is evaporated 
onto the PMMA, the result is very poor and rough as I believe the silicon 
must be at too high a temperature for the PMMA, even after being baked at 
190 celsius overnight.

Therefore, my second question is that, does anyone know of an organic 
ploymer that can be easily spun to about 200nm thickness onto a silica 
substrate, and furthermore, that can take the evaporation of such a high 
temperature as silicon (perhaps after baking) and yet will be can be 
relatively easily removed afterwards in a wet etch that will not attack 
metals.

Or, if there is nothing such as this, does anyone know of a suitable 
alternative barrier layer that will either not be etched by a pure O2 plasma 
and can be etched either at the same rate as photoresist if O2 is necessary 
or with a plasma without O2 present. It must also be able to be deposited 
(evaporated?) without damaging the underlying base layer of either PR or 
PMMA.

Thanks you so much for any help,

Martyn Gadsdon