Particle Problem - or - Randy True is the SNF Labmember-of-the-month!

Mary Tang mtang at
Thu Sep 29 14:09:49 PDT 2005

Hi all --

A particle problem was reported which is now suspected to have 
originated in the spin rinse dryer at wbnonmetal.  Randy's wafers and 
Jasmine's wafers were badly affected by this.  They had processed 
material there on Tuesday afternoon.

We strongly suspect that a wafer had been broken inside the spin rinse 
dryer, but cleaned itself after many cycles.  A spin rinse dryer runs at 
very high speeds so that a broken wafer will become pulverized into 
silicon dust.  Wafers processed there afterward will become contaminated 
for several cycles, until the chamber is flushed clean.  The 
contamination that Randy describes (very fine, crystalline particles on 
the surface, varying in density from wafer to wafer and across some 
wafers, but spread out, not patchy) is very consistent with silicon dust 
in the spin rinse dryer.

So...  Let's make sure this doesn't happen again.  If your wafer breaks 
in the spin rinse dryer, shut the dryer down, make a comment on Coral, 
and notify a staff member.  Don't be afraid to tell someone -- accidents 
happen and wafers do break.  Don't just wipe it up and let it go -- it 
needs to be thoroughly flushed, cycled several times, and then tested.  
It's not necessary to shutdown the wbnonmetal bench (although you should 
do this if in doubt), because people can still dry their wafers at 
wbnitride (you don't need to enable to use a spin rinse dryer.)  But 
notifying a staff member and putting a note on the spin rinse dryer is 
something anyone can do.  *Please realize that if you don't do this, you 
have contaminated other people's wafers. *

Please also understand that wafers are subject to a lot of force in the 
spin rinse dryer.  Think twice before putting wafers into the spin rinse 
dryer which may run the risk of breakage. (wafers with chips or damage 
on the edges, wafers with deep trenches following the <100> crystal 
planes, etc.)

If you have processed wafers through wbnonmetal some time Tuesday and 
maybe even Wednesday, you should inspect your wafers for particles.

Finally, I would like to add that Randy has been, absolutely, the model 
labmember through all of this.  First, he was doing all the "good 
engineering practices" that we should all try to do:  he began with a 
lot of extra wafers and split them up to minimize risk, he added in test 
wafers at each step, and he visually inspected after each critical step 
in the process.  Somewhere between his inspections before P5000 etch and 
after wbnonmetal clean, he found the particles on his wafers.  As soon 
as he discovered this, he made comments on Coral for these tools and 
notified staff members the following day.  Having isolated the suspected 
steps, he then ran some more test wafers and verified that the P5000, 
the gasonics, and the wbnonmetal hot pots and rinse tanks were fine.  
Many thanks to Randy for following up on this problem and working with 
us on it!


Mary X. Tang, Ph.D.
Stanford Nanofabrication Facility
CIS Room 136, Mail Code 4070
Stanford, CA  94305
mtang at

More information about the wbmetal mailing list