residue on wafer

Elizabeth Edwards ehe at
Wed May 14 12:28:06 PDT 2008


The only think I can think of is that fine CMP particles stuck in
microscopic cracks or surface porosity. You might have gotten a more
textured run than normal due to the bad TEOS. The reason is that even
after lots of cleaning I see what I'd guess to be CMP particles lodged
in fine cracks on the polished surface (SEM images attached).


On Fri, May 9, 2008 at 6:52 PM, Christopher McGuinness
<cmcg at> wrote:
> Hi all,
> I have recently noticed an odd residue on my wafers.  It looks like a big
> animal sneezed on them.  I am suspicious that this came from either the CMP
> or the TEOS tubes, and was wondering if anyone noticed anything similar in
> the past.  My observations are as follows:
> CMP: typically leaves a slight hint of slurry residue if not cleaned
> properly afterwards.  At this point I am fairly confident in my cleaning
> procedure, and have not had a problem with this on the last batch of wafers
> I ran, so i'm somewhat doubtful this is the culprit.
> TEOS: Tube ran out of TEOS during an earlier deposition.  I ran several
> tests after this point, and reran the wafers to finish the oxide deposition.
> At this point I do recall noticing a slight hint of something, but it was
> not obvious enough to cause alarm.  After etching into this layer the
> residue was more apparent, and after depositing a layer of poly on top of
> this it was very apparent.
> I have included a picture (as best of one as Uli and I could get).  Anyone
> see anything like this before, or think this will cause serious problems?
> Any ideas of what it could be?
> Thanks,
> Chris McGuinness
-------------- next part --------------
A non-text attachment was scrubbed...
Name: ehe_CMP_crackresidue_zoomedout.JPG
Type: image/jpeg
Size: 52638 bytes
Desc: not available
URL: <>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: ehe_CMP_crackresidue_zoomedin.JPG
Type: image/jpeg
Size: 22474 bytes
Desc: not available
URL: <>

More information about the cmp mailing list