PQuest update 7-8-08, jimkruger
jimkruger at yahoo.com
Tue Jul 8 17:18:29 PDT 2008
PQuest update 7-8-08
My belief is that there are two issues with PQuest at the moment.
Most serious is a varying air leak at the mechanical rotary feed-thru that raises and lowers the wafer clamp.
The feed-thru is FerroFluidic, using fluorocarbon based oil for service in the plasma chamber. Repair is likely not feasible. Estimated delay for a replacement part is 4 weeks (from today, order placed).
The second problem is a varying leak across the gate valve when closed. This does not affect the system in normal operation but interferes with leak-up tests used to measure the feed-thru leak. If the leak is significant, the pressure “zero” will be affected so that very low pressure recipes (GaAs is at 2 mTorr) may run with a significant pressure error.
The more serious problem is the addition of an uncertain flow of air to the recipe. Recent experience with GaAs has reported rapid loss of resist as well as inability to strip some areas of resist (encapsulated by “GaAs oxide” or perhaps Boron Oxide from the BCl3?).
I recommend that recipes not involving a significant deliberate flow of Oxygen not be run until repairs can be done. Even a qualification run may not be useful since the leak appears to vary with every motion of the wafer-lift clamp.
For the record, the historical “good” leak-up rate is 1 to 2 mTorr/minute, measured over 3 to 5 minutes. Currently, I see as much a 10 times that, but with high variation due to the varying leak across the "closed" gate valve.
A test recipe demonstrating the 2 issues is “_TST1”.
Bottom line: expect about a 4 week delay before Oxygen free etching can be restored.
More information about the pquest