[labnetwork] pmma cracking issue

[John Watson - TNW]

Hi all,
Thank you all very much for all the responses; this gives us some good ideas of things to check into.  In response to a few of the questions that were raised:

1) We have independently checked the hot plate temperatures, and they are a bit low (by ~15C), so we can try bumping the setpoints up a bit to compensate.
2) The polysulfide treatment is certainly the least standard part of our process, though it is very weak (diluted 200:1).  We are planning on adding control samples without the polysulfide soak when we do our process runs to see how much difference this makes.
3) Heating during deposition/He etch.  The samples are clipped to a 4" x 3/8" copper plate.  After the deposition the copper is probably ~30-35C; I'm not sure if the resist surface could be hot enough to crack in this case, but we will start comparing the emission current from good and bad runs and check on the stability of the cooling water to the hearth.  The gold crucible (graphite) glows bright orange during the deposition (so far I have not found any sweep parameters that can keep it cooler), so that probably doesn't help the situation. 
4) State of resist at various stages of the process. We have inspected the resist after development and don't see any cracks.  Most of the time we don't check the resist after the polysulfide since we want to avoid prolonged air exposure, but we will check more into this.  Attached is an optical image of a bad chip before liftoff; given that the metal hasn't started to wrinkle I would guess the stress is not extremely high, but maybe it is high enough...
5) Stray exposure/faulty blanking.  We have seen the cracks with devices patterned in three different e-beam systems (two Vistec, one Raith with the Raith using different file formats), so this along with the fact that we don't see the cracks after developing would seem to rule out extra exposure in the EBPG.  

Thank you all again,

John




-----Original Message-----
From: labnetwork-bounces at mtl.mit.edu [mailto:labnetwork-bounces at mtl.mit.edu] On Behalf Of Mark K Mondol
Sent: Tuesday, November 17, 2015 10:26 PM
To: labnetwork at mtl.mit.edu
Subject: Re: [labnetwork] pmma cracking issue

John:

Thanks for documenting the process so well.

I am not sure that I would put this down to resist cracking, although it might be. My first thought was a metal evaporation issue, but you say it happens with different evaporators and sputtering systems (impressive to liftoff what look to be 100nm lines with a sputtering system putting down the metal).  Stress in the metal seems more likely than stress in the resist. Almost any hotplate bake >160C should relieve stress in the PMMA. My second thought, uniformed by any real knowledge of chemistry, is that your Ammonium polysulfide is too base. In my experience base solutions attack PMMA and ruin it as a resist.

The other, obvious, culprit is the He Ion Milling as it is common to all your failures; though I can't think of a particular mechanism that would create this fault.

In many years of ebeam lithography I have never found PMMA resist to be the problem (i.e. it is extraordinarily stable) other than when it was inadvertently exposed to x-rays. So I would not be looking at the initial resist.  Baking PMMA is also a very tolerant procedure. On the other hand your resist profiles must be suffering, to some extent, by the exposure to heat (plasma ash and possibly He ion milling and possibly by the metalization.  The doses you mention and development times and developer all seem very reasonable.

Have any users examined the exposed and developed PMMA prior to metalization?  I think that would be my first step.

Regards,

Mark K Mondol


--
Mark K Mondol
Assistant Director NanoStructures Laboratory And Facility Manager Scanning Electron Beam Lithography Facility Bldg 38  Room 177 www.rle.mit.edu/sebl mondol at mit.edu office - 617-253-9617 cell - 617-224-8756


_______________________________________________
labnetwork mailing list
labnetwork at mtl.mit.edu
https://www-mtl.mit.edu/mailman/listinfo.cgi/labnetwork
-------------- next part --------------
A non-text attachment was scrubbed...
Name: before liftoff.jpg
Type: image/jpeg
Size: 157066 bytes
Desc: before liftoff.jpg
URL: <https://mtl.mit.edu/pipermail/labnetwork/attachments/20151118/e52ec96d/attachment.jpg>