Dear all,
I am Isaac Chan from the University of Waterloo in Ontario, Canada.
Our research group have a Trion ICP/RIE etcher for developing our dry etch
process for thin film transistor. We are currently encountering problems
with photoresist ashing/stripping after commonly-used dry etch process.
After dry etching, the photoresist cannot be stripped completely by solvent
stripper or ashed completely by oxygen plasma. Therefore, I need your
advice/knowledge on the possible cause and solution to this photoresist
residue problem.
I would really like to know what this photoresist residue is. Is it the
so-called carbonized photoresist residue? If so, what causes it and how to
remove it? Also, is there any pre-etch treatment I can do to avoid this
residue formation. I heard that in an industrial standard dry etch process,
the substrate should be cooled during etching. Is this technique for
avoiding photoresist residue formation? Another uncertainty is whether or
not the chamber wall and electrode materials would have anything to cause
this issue.
Here is our experimental details for your information:
We test out this photoresist residue source systematically with different
gases (CF4/H2, CF4, H2, SF6, O2). The results of each set of dry etch
conditions are illustrated in the attached pictures.** We are using AZ3312
positive photoresist and sputter-deposited Cr on silicon wafer for this
experiment. The Cr film acts as a perfect etch-stop layer for all the gases we
used. So the patterns you see are photoresist residue patterns, not etched
patterns. It is surprising that even a fresh (un-processed) photoresist cannot
be completely removed by pure oxygen plasma (Pic.5). These residues are
loosely adhere to substrate, which can rubbed away in some degree (Pic.6) but
not entirely without damaging the substrate. They cannot be washed away
either. And in another process conditions (CF4/H2), the photoresist surface
after dry etching becomes rough (cannot be resolved in Pic.7 but is observable
under optical microscope). Then after immersing the substrate into hot stripper
bath for more than 15 mins, the photoresist structure seems to be "collapsed"
(Pic.8). With a scratch on this residue, a "skin" of photoresist is scratched
off
(Pic.9). This suggest the stripper can only strip the "good" photoresist under
this insoluble "skin" of photoresist, which had exposed to dry etching. Similar
results are also observed in different models of photoresist (AZ4330 and SU-8)
on glass and silicon substrates. All these results suggest that the residue is
probably not related to a particular gas or photoresist model, but something
else more fundamental with photoresist interaction in dry etch environment.
Thank you very much for your time. Any suggestions or advices will be
greatly appreciately and thank you in advance!
Yours sincerely,
Isaac Chan
Ph.D. candidate in a-SiDIC research group
Electrical and Computer Engineering
University of Waterloo
200 University Ave. W
Waterloo, Ontario, Canada
N2L 3G1
(529) 888-4567 ext.6014
**File list:
1. Cr on c-Si etch pattern - AZ3312PR CF4-20%H2=16-4sccm ICP=300W RIE=25W 10mT
12min(overetch).jpg
2. Cr on c-Si etch pattern - AZ3312PR CF4=20sccm ICP=300W RIE=25W 10mT
10min(overetch).jpg
3. Cr on c-Si etch pattern - AZ3312PR H2=20sccm ICP=300W RIE=25W 10mT
15min(overetch).jpg
4. Cr on c-Si etch pattern - AZ3312PR SF6=20sccm ICP=300W RIE=25W 10mT
12min(overetch).jpg
5. Cr on c-Si etch pattern - AZ3312PR O2=20sccm ICP=400W RIE=25W 10mT
3min(overetch).jpg
6. Rubbed residue pattern, Cr on c-Si - AZ3312PR CF4-20%H2=16-4sccm ICP=300W
RIE=25W 10mT 12min(overetch).jpg
7. Cr on c-Si etch pattern before strip - AZ3312PR CF4-20%H2=16-4sccm ICP=300W
RIE=25W 10mT 5min(underetch).jpg
8. Cr on c-Si etch pattern after strip - AZ3312PR CF4-20%H2=16-4sccm ICP=300W
RIE=25W 10mT 5min(underetch).jpg
9. Cr on c-Si etch pattern after strip with scratch - AZ3312PR
CF4-20%H2=16-4sccm ICP=300W RIE=25W 10mT 5min(underetch).jpg
(If the pics does not display on the mail list, I can send them to you
directly.)
