Request for methanol based rinse following HF dip in wbdiff

Karthik Vijayraghavan karthikv at
Mon Feb 1 22:33:08 PST 2010


I have a process step where I have an SOI wafer on which I intend to deposit
silicon nitride in tylannitride. My process is expected to be sensitive to
the native oxide that is expected to be present between the silicon wafer
and nitride layer above it. I would like to use a slightly modified diff
clean that differs only in the way the wafer is rinsed following the final
HF dip.

Standard diffusion clean in wbdiff involves an HF dip which is meant to
remove native oxide from the surface. However, this native oxide quickly
grows back by the time the wafer is rinsed in water and transferred to the
respective furnace. I would like to use a methanol based (4:1 methanol:water
followed by 100% methanol) rinsing solution instead of DI water. This
treatment leaves the surface with Hydrogen termination which in turn
suppresses the growth of native oxide long enough to transfer the wafer into
furnaces. I will use my own clean recommended labware and will follow any
cleaning procedure required prior to putting the wafers inside the furnaces.

Please find attached work from Prof. Howe group (when he was at Berkeley)
which shows how this method works. Specifically look at "Treatment I" in
"Table I" and the results in "Table III".

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: Houston - Effect of hydrogen on adhesion - 1997.PDF
Type: application/pdf
Size: 571218 bytes
Desc: not available
URL: <>

More information about the specmat mailing list