[labnetwork] Bosch etch foreline and pump powder residue question...

[Aebersold,Julia W.]

Holy flying shrapnel!  Glad the system contained the flying bits.  That could have been very bad if not contained and went into the facility or hit personnel.

Cheers!

Julia Aebersold, Ph.D.
MNTC Cleanroom Manager
Shumaker Research Building, Room 233
2210 South Brook Street
University of Louisville
Louisville, KY  40292

502-852-1572
http://louisville.edu/micronano/

From: Ian Harvey [mailto:IRHarvey at eng.utah.edu]
Sent: Tuesday, November 5, 2013 5:20 PM
To: John Shott; Aebersold,Julia W.
Cc: labnetwork at mtl.mit.edu
Subject: Fwd: [labnetwork] Bosch etch foreline and pump powder residue question...

Dear John, Julia and all respondents:

Thank you all for your helpful feedback.  We now have a reasonable procedure to use when the refurbished pump and bonnet valve kit both arrive.  Summarizing from what we learned and from the collective responses:  PPE: disposable tyvek suit + C-filtered respirator, treating any volatiles as nasty - perhaps carcinogenic, and changing gloves often.   Treating wipes and gloves as hazardous waste.  Snorkel exhaust and cleanroom vacuum for dust, foil capping and bagging of the components, and brass wire brush / solvent+wipes / LN2 for foreline cleaning in the fume hood.  We will document the shipping materials for the returned pump with the source gases used (SF6, C4F8) and include the EDS analysis (also found out that Oerlikon is familiar with the residue and accepts this back properly capped).  Going forward:  Make sure all the heater elements are working, along with temperature interlocks, and keep your eyes on it for periodic maintenance.  Also:  PM the system or budget for pump replacement before the thing blows up and the core return has no value.

FYI: Among the responses, I received verification for why it is so important to pay attention to such a large pump going out of balance with caked-on materials:
"In the end it destroyed a $40K maglev turbo threw parts into the main chamber and the whole way back to the rough pump 25 ft away. It also bent a 2 inch thick stainless steel throttle valve and sheared off 4  ½'" thick stainless steel bolts, and dented the side of the system from the force of the abrupt stop.  The only part that I kept is what remains of the central portion of the pump that had the veins (once) attached.   Here are two pictures of how catastrophic it was. The first one (3480) is what it looked like when we unhooked the turbo. We didn't just dump parts into the containment vessel for the turbo that is how it came apart. The other one you can see the sheared off bolts and where the clamps grabbed the gate valve. Too bad I didn't get a picture of how bent it was. Enjoy and I hope you can fix your system before that happens." --Dave

Dave sent some fascinating photos of his pump, which casing thankfully contained all the schrapnel, and which looks like metal-noodle soup.

Labnetwork is a great resource and thank you again to colleagues who make this job so much easier because of your freely shared knowledge and experience.

-Ian

Begin forwarded message:

From: John Shott <shott at stanford.edu<mailto:shott at stanford.edu>>
Subject: Re: [labnetwork] Bosch etch foreline and pump powder residue question...
Date: November 5, 2013 9:24:21 AM MST
To: Ian Harvey <IRHarvey at eng.utah.edu<mailto:IRHarvey at eng.utah.edu>>
Cc: labnetwork at mtl.mit.edu<mailto:labnetwork at mtl.mit.edu>

Ian:

While you will likely get more thorough responses from others on this matter, let me share some or our experiences in this area.  In particular, these tools generate a lot of polymer.  The same chemistry that you are relying on generating the polymer on your wafers during half of the switched etch/deposition process is also going to produce a low-grade polymer everywhere else.  Much of the polymer that you find in the foreline tends to be rather gummy and difficult to remove ... at least that is our experience.  Many (but not all) machines have heated turbo pumps and I believe that had been found to be pretty effective in terms of minimizing polymer build up in one of your most expensive components on these tools.  One of our machines has a water cooled trap between the heated turbo and the remainder of the foreline ... but, to be honest, that trap has not proven to be particularly effective in our case.

Because the forelines tend to get pretty well coated with a sticky, gummy polymer, we have found them to be pretty difficult to clean.  One "trick" that seems to work pretty well in the case of the forelines is to very carefully pour some liquid nitrogen down the foreline which will freeze it and cause it to fall off in chunks.  This, of course, has to be done VERY carefully due to the hazards of working with liquid nitrogen including worrying about "burns" due to the extreme cold and asphyxiation due to displacement of oxygen in the air.  One thing that we haven't tried, but may be easier than dealing with liquid nitrogen is to see whether packing a length of foreline in dry ice may have the same effect ... dry ice has it's own problems because of it's very low temperature and sublimation of something that can displace oxygen, but it is probably easier to handle than liquid nitrogen.

While the polymer itself isn't wildly toxic or flammable in the way that some of your pumps and forelines on other deposition tools are (things that pump silane and DCS, for example) we treat all forelines very carefully and, in particular, for anything that is the least bit "powdery or dusty" make sure that we have adequate protection to avoid inhaling or ingesting anything coming from these pump lines.

Finally, depending on the distance your foreline pump lives from the main tool, you may need to consider upsizing the diameter of your foreline a bit to help offset the constriction that will take place as your get polymer build up between periodic cleaning.

Good luck,

John


On 11/4/2013 3:36 PM, Ian Harvey wrote:
Dear Labnetwork colleagues,

We are installing a used STS DRIE tool acquired from the used tool market and used for Bosch-Si etch.

We are unsettled by the large amount of white powder collected in the foreline and inside the large turbopump.  The turbo is becoming unbalanced as flakes come off, and we are getting ready to trade in the large maglev pump for a refurb, and clean the forelines.

Does anyone have experience with what the powder is (F-C-S-O / EDS analysis attached), its risks, and how to safely manage it during pump removal, packing and declaration for shipping?

Thanks!

Ian

********************************************
Ian R. Harvey, Ph.D.
Associate Director, Utah Nanofab
 &
Micron Microscopy Core
2511 SMBB (USTAR)
University of Utah
801/585-6162 (voicemail)
801/581-5676 (lab main number)
www.nanofab.utah.edu<http://www.nanofab.utah.edu/>


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://mtl.mit.edu/pipermail/labnetwork/attachments/20131106/8a039966/attachment.html>