I am not sure the exact reason why it roughens,
see below for a guess, but I can tell you it is
not the residue of etching byproduct since the
roughness is easily removed with acetone/ipa
(that and the byproducts of a Sf6 O2 etch are all
gaseous for Si, Si3N4 or SiO2, which is the
reason it is so common).

Nicolas Duarte

My Guess-I have a feeling that it has to do with
a combination of crosslinking due to the RIE
process (making some sections of the PR harder to
remove) and bombardment of other sections,
causing very small roughness at first and then
giving way to cumulative effects, but that's just
my guess

At 11:46 PM +0800 12/13/05, =?gb2312?B?x8e089PC?= wrote:
>thanks for your explanation, yes, the bombardment of ion will decrease the
>thickness of the phtoresist, but the problem confused me is why the surface of
>remained photoresis is not smooth, but roughed? Is the residue byproduct of
>etching responsible for this?