I agree with Bill. Silicon nitride is highly variable.
Here are a few more comments:

PECVD nitride tends to be tensile when deposited at 13.56 MHz,
but compressive (due to ion bombardment) when deposited at 380 kHz.
Both of these frequencies are available on the same system from STS.
By alternately depositing at low and high frequency for few seconds each,
a lower-stress film can be produced.

I have found that increasing the ratio of SiH4 to N2 (I used N2 rather than NH3
as the nitrogen source) can also change the stress from tensile to compressive.

Furthermore, thermal processing can change the stress. As PEVCD nitride
films typically contain a lot of hydrogen (>10%, according to one reference),
the hydrogen can be outgassed at high temperature.
In our experiments, this moved the stress toward more tensile.

        --Kirt Williams Agilent Technologies


> -----Original Message-----
> From: beaton@npphotonics (Bill Eaton) [mailto:[email protected]]
> Sent: Friday, February 28, 2003 9:40 AM
> To: [email protected]; 'General MEMS discussion'
> Subject: RE: [mems-talk] Residual Stress of Silicon Nitride
>
>
> Stress in nitride films is an interesting, and sometimes
> maddening topic.
> Stress in a stoichiometric film (i.e. Si3N4) is very high,
> typically 1 GPa
> tensile. Because of this huge stress, film thicknesses are
> limited: wafers
> will break or be severely warped if the thickness is to large.
>
> Usually non-stoichiometric films are deposited to combat this
> issue. These
> films are silicon-rich -- closer to Si4N4 than stoichiometric
> Si3N4. Film
> stress tend to be in the range of 10-100 MPa. Depending on
> your application,
> you have a couple of options, LPCVD or PECVD. LPCVD films
> will tend to be
> pinhole free and are (generally) more resistant to wet
> chemical attack.
> However LPCVD films also tend to be non-uniform, with
> thickness and stress
> variations from the edge of the wafer to the center of the
> wafer and also
> from the top of the boat to the bottom of the boat.
>
> PECVD films are more uniform (I think). Historically, PECVD
> nitride wasn't
> suitable for a KOH etch mask, because of excessive pinholes.
>
> Bill Eaton, Ph.D.
> Materials & Analysis Manager
> NP Photonics
> mailto://[email protected]
> www.parvenutech.com
>
> > -----Original Message-----
> >
> > I really appreciate if you could help me find data on
> > residual stresses
> > left in Silicon Nitride layers on usual and hi-quality
> > wafers. For our
> > research, we would need to find very-low-stress
> > nitride-layered wafers or
> > methods to create them by our own. Prices, availability and
> > literature are
> > welcome.
> >
>
> http://mail.mems-exchange.org/mailman/listinfo/mems-talk
> Hosted by the MEMS Exchange, providers of MEMS processing services.
> Visit us at http://www.memsnet.org/
>
>
> _______________________________________________
> [email protected] mailing list: to unsubscribe or change your list
> options, visit
http://mail.mems-exchange.org/mailman/listinfo/mems-talk
Hosted by the MEMS Exchange, providers of MEMS processing services.
Visit us at http://www.memsnet.org/