Advice on Through Wafer Etching

Generally, you should be fine with around 1um of oxide, especially if you
use thermal oxide.  Alternatively, use metal.  The drawback with oxide is
that you can get lateral etching (notching) at the oxide interface unless
you take special precautions.  This is not seen with metal, as it is caused
by a charging effect.
The helium would be at a nominal value of between 5 and 10Torr, but the true
value behind the wafer is likely to be less than that.  Very thin layers
could break because of the helium pressure, but 1um should be sufficient
unless you are doing a very long over-etch.
You are right to be concerned about making holes through the wafer.  Once
one point breaks, the helium pressure behind the wafer drops and the thermal
path to the electrode degrades, causing the etching conditions to change
significantly.  This can either mean that you never finish etching some
parts, or you get dramatic profile changes.

Regards,

Martin Walker, Applications Engineer, Oxford Instruments Plasma Technology
(http://www.oxford-instruments.com/plmchp5.htm)
----- Original Message -----
From: "Sun Yu" 
To: 
Sent: Friday, November 16, 2001 2:18 AM
Subject: [mems-talk] DRIE through etching


> Dear friends,
>
> Have you etch through the whole wafer in a deep trench etcher without
using
> a dummy wafer as a carrier below the device wafer? If you used SiO2 layer
> as etch stop, how thick did you make your SiO2 layer? Is it possible that
a
> SiO2 layer of 1um might explode right before etching is through because of
> the Helium pressure from the chuck? Thank you very much.
>
> Best,
>
> Sun
> _______________________________________________
> mems-talk@memsnet.org mailing list: to unsubscribe or change your list
> options, visit http://fab.mems-exchange.org/mailman/listinfo/mems-talk
> Hosted by the MEMS Exchange, providers of MEMS processing services.
> Visit us at http://www.mems-exchange.org/