Problem lampoly SNF 2010-12-10 16:41:55: Conditionally up

mtang at snf.stanford.edu mtang at snf.stanford.edu
Fri Dec 10 16:41:56 PST 2010


Dear lampoly users:
The machine is conditionally up.  It tunes beautifully now, thanks to expert attention from Jim McVittie. A couple of issues to note, however:
1.  DC bias is lower than before.  Increasing RF power to the bottom electrode doesn't increase DC bias.  General etch qual results look good(see below); however, profile has not yet been confirmed with SEM.   Please check before committing your critical wafers.
2.  Jim observes that HBr/O2 only (no chlorine) recipes, such as #25, do not tune quickly and even exhibit some plasma instability for 30 seconds.  This is not a machine problem, but one of the chemistry.  Jim has shown that 10 seconds of a standard recipe 1 main etch (step 5) allows very fast tuning/stable plasma for HBr/O2 recipes.  This is something we should look into for people using these recipes.
3.  The receiver cassette wafer present sensor sometimes fails, particularly on some wafers with certain backside films.  This needs to be adjusted or replaced, but in the meantime, please make sure wafers unload into the receiver cassette -- watch your wafers!  If a wafer is not sensed, it can go back into the exit loadlock and back into the chamber.  This could result in damaged/broken wafers.
Etch rate results (resist on silicon, etched on Recipe 1):
Silicon etch depth = 3985A
%nonunifomity = 1.3%
Silicon to resist etch selectivity = 4.5X
Everyone:  please share your observations with lampoly etch users.
Thanks.




More information about the lampoly-pcs mailing list