[labnetwork] Problems with ALD system

[Hathaway, Malcolm]

Hi David,

This is Mac Hathaway, at Harvard CNS.  I'm not familiar with A155, but the oil should be of a sort that is inert and appropriate for O2 service.  This usually mean Fomblin, or some other perfluorinated oil of that type.  If you go with fomblin, you have to make sure the pump seals are of the appropriate type, as you mentioned.

To provide a place for our excess precursors to "wet out" and react before they reach our pump, we put a plain old mist filter on our Cambridge Nano/Ultratech Savannah unit.  I should think you could do something similar and have a positive effect on your backstreaming.  If the Opal has pressure monitoring, the pressure spikes associated with precursor pulses will be softened by the presence of the mist filter.

Have you done any XPS on your film to see if there is any carbon/pump oil in it?

"Dirty" wafers would presumably have nucleation issues, but if you really see copious oil in your chamber, that would seem to be the most likely cause of dirty wafers...

Does the Opal have a constant N2 purge?

Last thing (or perhaps first thing!):  Check with Oxford to see what they recommend.  If your unit is stock, they will almost certainly have seen this before, and have good suggestions...


Mac
________________________________
From: labnetwork-bounces at mtl.mit.edu [labnetwork-bounces at mtl.mit.edu] on behalf of David Frankel [frankel at maine.edu]
Sent: Friday, April 18, 2014 1:25 PM
To: labnetwork at mtl.mit.edu
Subject: [labnetwork] Problems with ALD system

We have an Oxford Opal ALD system we use primarily for Al2O3 deposition using TMA. It is a thermal system, no plasma.  It is pumped by a rotary vane pump with a synthetic di-ester oil (A155), same as originally supplied by Oxford. There is no trap between the pump and chamber. We are having problems getting consistent good film quality.  We often see films with poor quality, micron-sized pin holes etc...  We think the problem may be related to contamination of the chamber with back-streaming pump oil as we often find considerable quantities of oil in the chamber.
Some questions:
1) The A155 oil has a fairly high vapor pressure, is there a lower vapor pressure oil you can suggest that would stand up to the process gas that would work with our present pump?  (switching to a dry pump is cost prohibitive at this time, switching to Fomblin may be possible but would also presumably be expensive as it may require a pump rebuild or exchange)
2) Can a molecular sieve trap be added in the fore-line without creating any further problems?  If so, can a standard one be used or must it have some special features (high conductance...)?  Are there other types of traps that would work in this application?
3) With a trap in the line what safety measures are needed, for instance when venting to air?
4) Are there other issues that could be causing poor film quality?

Thanks for your help on this,
David

--
David J. Frankel - Senior Research Scientist
LASST - University of Maine
5708 Barrows Hall
Orono, ME 04469-5708
207-581-2256
frankel at maine.edu<mailto:frankel at maine.edu>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://mtl.mit.edu/pipermail/labnetwork/attachments/20140418/221e9c5a/attachment.html>