Metalica problems

Eric Perozziello eap at gloworm.Stanford.EDU
Tue Jun 19 14:26:18 PDT 2001

Dear Metalica Users,

	Metalica has been getting "severe duty" usage levels since the 
innotec has taken a vacation, and I wanted to send out a message 
to make sure that we're all on the same page in terms of proper 
usage. You should know that many parts for this system are no longer
supported by the manufacturers, so if we break them, we're going to 
be without "dirty metals."

	The biggest problems we've had recently (since the innotec 
power supply "evaporated") have been limited time and very short
cryo regen cycles.  I think both of these problems can be avoided
with a little effort.  

	This machine is designed for low throughput, and generally 
thin metal films. It has very few interlocks, so unlike many other 
more modern machines in the lab, it is very easy for you to break it
(or it to break you!) if you're careless. It's not an automated machine.


	For the scheduling problems, we should recognize that the 
cryo (when operating at cryo temperatures) is running very well. 
Typipcal pumpdown into the minus 7's takes less than an hour after
opening the gate valve.  What this means, is that you only need about 
2 hours for an average run.  Pumping longer doesn't make prettier 
films.  So, when scheduling your runs, try to make the reserved time less than 
3 hours, especially during the day.  (Of course, if you have multiple runs,
it will take longer).  Also, try to minimize overnight pumpdowns during 
this heavy use time we're in.  If you must do an overnight pump, try to 
start it after 11 PM, and be out by 7AM. I really don't want the lab to 
put a user time limit on this machine like we have on the gryphon,
primarily because this system is designed to be more flexible. Let's 
fix the scheduling problems ourselves so they don't have to!

	The problem with very short cryo regen cycles is two-fold. First 
of all, the occasional mistake happens-  someone opens the vent with the 
cryo still open, or worse, opens the roughing valve and cryo gate valve
simultaneously.  This isn't the end of the world, but does take the system
down for a day or two.  The other problem is less clear to me.  Even under 
heavy usage, the cryo should last longer than a week and a half.  Maybe 
someone is doing extremely long sputter runs, or turning up the argon flow (?).
Or someone might be turning on the argon valve while the MFC is off (you 
shouldn't allow this condition for more than a second or two, since the MFC
is fully open when it's not turned on.) If you know something about this 
that I don't, please let Jim Hayden or me know!  


	Another problem is that I keep finding broken pieces in the 
chamber.  While I think everyone who has used this machine a lot has 
broken a wafer in there, it seems now that someone's wafers have dandruff.
The problem that we're seeing is that these pieces get down into the gate
valve and cryo pump, which may be causing some of our short regen problems.
and also may contribute to some of the "arcing" problems that have been 
reported. If your wafers are breaking when you press "Bell Jar Up" then 
please ask me to show you a more gentle way to do this.


	There are other things, which seem to be less frequent: 

* Target overheating- Metalica is a low throughput research system
that was put together here primarily by Chris Storment. The sources
are max 100W.  (Just for reference, the Gryphon is 7500W, or 75 times
higher, and the innotec, though an evaporator, was 15000W). Metalica 
only holds 3 wafers plus a dummy.  Try to plan your process accordingly.
Thick films on lots of wafers is probably a bad idea if you plan on 
using metalica. Use 50W max for long runs (more than 20 min per source)
or pause to let it cool.  Never use over 100W in any circumstance.
* Targets disappearing-  The lab provides non-precious metal targets
for general use.  Some of these have disappeared.  This means that the
next user will lose their run if they counted on one being available. If 
you need to have your own target, ask me and I'll tell you where to get 
them.  They're generally not that expensive. One user had her personal 
high-purity gold target stolen from her storage box.   It would really be 
nice to not have to keep an armed guard in the lab to store targets there! 

* Overtightening screws-  The screws on the target clamp and anode are
really just "locator pins" and do not need to be cranked in tight. Barely
finger tight is plenty of torque.  If you overtorque them, they can vacuum 
weld in place.  

* Pump purge- I think we've taken care of this problem by removing the 
valve on the pump purge line.  

* Burning through a target-  If in doubt, don't use a marginal target for 
a long run!  If the tagert burns through, it can damage the machine. 
The parts we'd have to replace are no longer made.  Be careful.

* "Arcing"-  The plasma seems to strike, then go out, then strike......
I've found one instance where this was partly due to improper target 


	Please don't hesitate to talk to Jim Hayden or me.  We'd all 
rather have you bug us with a question than to have you damage the machine!  
Please stick to the standard procedures, or ask before trying anything 

Happy Sputtering,

Trivia Question: Who named this machine, and why did they choose this
Prize:  A Gold Dummy Wafer that you can frame and hang prominently on the wall.

More information about the metalica mailing list