Hi Lihuan,

Thanks for the additional info - this helps. It would appear that you're
left with running a SF6/O2 plasma for both etches. For simplicity I would
suggest
running almost identical conditions as starting points and optimize from
there.

Reasonable starting points would be (making some assumptions on
configuration
as this will impact power settings!

I would siggest a 10% O2 in SF6 gas source with flows such that you get a
stable
APC at about 10mT. Coil power at whatever you know runs a stable discharge
and then adjust platen power to approx 80V or so for the SiN and 200V+ for
the
SiO.

The vendor should, of course, be able to provide better info than this but
this would
be a reasonable start if you do not get info.

Good luck, Andy

Message text written by General MEMS discussion
>
Hi, Andy,

My structures are pretty big, so the profile requirement is vitually not a
constraints (CD > 5 µm and film thickness of both SiN and SiO are < 600 nm
that is also the depth to be etched, again, don't worry about the
selectivity.  Resist > 3 µm thick).  Anisotropy is not very important
either
for the same reason.  Basically I just need something works that could give
me
a satisfactory etch rate to avoid the burning and baking of resist too
much.
The substrate would be cooled by He backing which should be efficient I
guess.  I haven't contacted the vendor yet but will do so in a moment.
Thank
you for your time.

Lihuan

Andy McQuarrie wrote:

> Hi Lihuan,
>
> Both SiO2 and Si3N4 can be etched with SF6 as the F source. However,
> it's impractical to suggest starting conditions without knowing
additional
> constraints such as degree of anisotropy, profile requirements, etc.
> - Nominally, SiN will etch in low ion energy regimes whereas SiO requires
> higher ion energies - however, this also tends to be reflected in the
etch
> profile seen in F plasmas without passivation components (SiN often
> will be isotropicly etched).
>
> Therefore additional information that would help would be:
>
> - Profile and CD requirements
> - Film thicknesses (SiN, SiO and resist)
> - Whether substrates are actively cooled (or whether you're using small
> parts of a wafer on a carrier).
>
> Apologies for answering a question with a question (a few questions that
> is)
> but this will help in answering you. On a practical front, have you
> contacted
> the vendor?
>
> Good luck, Andy
>
> Message text written by General MEMS discussion
> >
> Dear experts,
>
> We only have the following gases available in our Oxford plasma lab ICP
> etching system, SiCl4, Cl2, CH4, H2, SF6, O2, Ar and N2.  I want to etch
> both SiO2 and Si3N4 (seperately, selectivey not needed, layers were CVD
> grown on III-V, say GaAs substrates) with a acceptable etch rate, say
> >50nm/min (preferably >100nm/min).  I prefer to use resist as mask.  Can
> anybody give me a receipe such that I have something to start
> optimzation with?  Thanks a lot.
>
> Best regards,
>
> Lihuan
>
> --
> Lihuan Song<