durusmail: mems-talk: post-CMP cleaning residues and issues
post-CMP cleaning residues and issues
2009-04-02
post-CMP cleaning residues and issues
my2232@columbia.edu
2009-04-02
Hello all,

I wonder if anyone of you can give me some advice about how to clean
my wafers after a CMP process.

I am etching oxide to level it with surface of the  silicon substrate,
in other words I am trying to level oxide and silicon at the same
plane. So, in the end of the CMP process I still have both oxide and
silicon on my wafer surface. In a more detailed explanation: I have
oxide trenches, and I want to level the surface of oxide with the
surface of silicon substrate. Initially oxide is at higher level
compared to the silicon substrate surface.

I have some residue issues after the post-CMP cleaning process that I
am following below:

Just after my wafer is processed in the CMP tool (Strasbaugh 6EC),
I quickly transfer it to a bucket filled with DI water (60 units) and
tetramethyl ammonium hydroxide (1 unit), and keep it there until I am
done with the remaining wafers.

I repeat this step to all of my wafers.

Then, when all the wafers are transferred into this bucket, I prepare
another bath composed of DI water (6 units), Ammonium Hydroxide (1
unit), Hydrogen Peroxide (1 unit). I increase the temperature of that
bath to 70 degrees Celcius, and transfer my all wafers into that
mixture and keep and agitate them inside that bath for 15 minutes.

Once that step is done, I am switching back to a fresh (or new) DI
water (60 units) and tetramethyl ammonium hydroxide (1 unit) mixture,
and using an ultra-sonication tool, I keep each of my wafers in
ultrasonicated bath for 5 minutes. After that step, I use a post-CMP
cleaning sponge, and I am literally washing my wafers by my hands by
rubbing them to that sponge.

I repeat this last step (washing by my hands) for 3 times (each time
with a new prepared mixture).

Last time, I stopped at that step and looked my samples under SEM. I
saw a level
difference between my thermal oxide and silicon layers (thermal oxide
is ~200nm higher than the silicon substrate...)

Then, I went on and kept my wafers in BOE(30:1) for 8 minutes, the
oxide and silicon surfaces were almost leveled, but other issues
occurred. But, still I have some serious amount of residue on my
wafers which affects the rest of my process.

If you know what is wrong with my post-CMP cleaning process, and share
that information with me I would be very glad because of your help.

Thanks in advance,

Best regards,

Mehmet Yilmaz
Mechanical Engineer
reply