Julie,
First of all, 10um is a very thick BOx. I've had problems getting more than
2um from vendors. I'm going to assume you have a very good vendor and that
the oxide is a simple Wet Ox. This may not be a good assumption though.
There is probably some kind of polymer residue left over from the DRIE on
the oxide. Before you go into the HF, piranha dip your wafers if you can,
or O2 plasma them if you have metals on them already. It's a good idea to
do this before any wet chemistry.
You shouldn't have to agitate your etchant. I think it's dangerous. If you
use 49% HF with a drop of Trition X-100 you shouldn't form any bubbles or
have any reason to use agitate. It also means that for a single wafer you
only need 2-3mm of chemical.
You may or may not know this, but the etch profile you get when etching the
BOx layer on an SOI is not isotropic. The latteral etch rate is 2-100 times
higher than expected at the interface of the two wafers. The increased
etching rate is constant for HF, BOE and vapor HF. With your 10um BOx that
number is probably very high. If you haven't already, you should look at
your etched wafers with an IR microscope to see what the implications are.
Gabriel
---------- Forwarded message ----------
> From: Julie Verstraeten
> To: General MEMS discussion
> Date: Tue, 2 May 2006 11:28:04 -0400
> Subject: [mems-talk] Problem in etching buried oxide
>
> Hi,
>
>
> I need to etch holes through SOI wafer (for front/back alignment). The
> holes
> have a diameter of 200 microns.
>
> The procedure is as follow:
>
> - first of all I deep etch (DRIE) the holes through the device layer (30
> microns
> thick) until I reach the buried oxide,
>
> - then I wet etch the oxide layer (10 microns thick), in the bottom of the
> etched holes (BOE),
>
> - at last I etch the holes (DRIE) through the handle layer (400 microns
> thick).
>
>
> The problem is that the buried oxide doesn't etch well.
>
> In the first 1 hour nothing happens as if there was a protective layer
> above the
> oxide. The etching of the handle layer however finishes by an etching
> period
> and there's no C4F8 plasma during this period to avoid passivating the
> buried
> oxide.
>
> After 1 hour it begins to etch (2-3 microns in two hours - BOE at
> 30°C, the
> liquid is magnetically agitated).
>
> After these 3 hours etching I started again 2 hours and nothing was
> etched.
>
> I don't know the nature of the buried oxide.
>
> Has anyone ever had the problem?
>
>
> Thank you,
>
> Julie