Etch the rectangular window around the metal line from the front in tmah. Then,
release the memberane using the Bosch process. Finally remove the  oxide using
hf vapor from the frontside.

On 24/10/2008, at 3:00 AM, "Taekyung Kim"  wrote:

Hello Folks,

I have a problem with SiO2/Si etching.

After Bosch etching from the back side of 400 um thick SOI wafer, 120 um x
120 um square membranes (SiO2 100 nm/Si device layer 3 um thick) were made.
Buried oxide layer was HF-vapor etched.
On the front side, I have a 10 um long, 250 nm wide Pt/Cr metal line and
want to suspend this metal beam by etching SiO2/Si.
Only 10 um long 5 um wide rectangle around the metal beam must be etched,
not the whole membrane area.

After electron beam lithography to define a etching window, the sample was
immersed in 1:6 BOE for 90 s (90 nm/min etch rate) and 25% TMAH at 70C for
12-13 min.
The problem is after TMAH etching, the whole 120 um square membrane was
damaged or etched maybe because once a window opens up on the bottom of Si
device layer, TMAH etches Si from the back side as well.

Any idea to control Si etching?
I desperately need to have SiO2/Si membrane with 10 um/5um rectangle
thru-etched window.