Weiquan,

The resist is carbonizing probably because it is getting too hot during the RIE.
Do you have a way to control the temperature of the wafer during the etch?  If
not, you can break up the etch into 5-min. long segments, allowing it cool in
between each etch.  Alternatively, try using BOE if you can tolerate the
undercut (roughly equivalent to the thickness of the SiO2 since it is
isotropic).  The etch rate is very controlled depending on the formulation.  You
could even etch 90% with BOE and do a final etch in the RIE to keep the
linewidth tolerance.

Brad Cantos
[email protected]
http://holage.com

On Feb 8, 2010, at 8:27 AM, weiquan yang wrote:

> Hi, everyone. I am trying to etch SiO2 layer by RIE dry etching using
> photoresist (s1827) as the etching mask. Because the SiO2 layer is as thick
> as 1 um, I need to do the RIE dry etching for long time. However when I
> extend the etching time more than 15 mins, the photoresist becomes
> carbonized. Carbonzied photoresist is hard to remove without attacking the
> GaAs device under the SiO2. Are there other SiO2 etching recipes
> without carbonizing the photoresist? Or are there some way to strip the
> carbonized photoresist without destroying the device? Thank you!
>
> Weiquan