Etching equipment for devices containing platinum

Jim McVittie mcvittie at snf.stanford.edu
Tue Jan 10 09:53:30 PST 2006


Lizzie,

Pt is a tough material to plasma etch. It does not form any volatile byproducts
at room temperature. One approach is to do the etch high temperature ( > 250 C).
See
http://www.tegal.com/pdfs/whitepapers/Etch_ISIF_2004_HTESC.pdf. Another approach
is to basically sputter etch it with Ar plus Cl2 and O2 at room temp. There are
a number of papers using this later approach with more modern etchers than the
MRC. See papers by Kwag-Ho Kwon from Korea  Yes, Kwon work does show Cl2 and O2
additives to Ar give some improvement in etch rate but there main effect is in
improved selectivity to an oxide mask. Kwon's increase in etch rate up to 150
nm/m is mainly from running at very high plasma densities.

While we have done some high temp (200C) etching in the PQuest, it is a pain to
do and I do not think we can go to a high temp to get a good Pt process. I
question whether Your SU-8 can take the temp. The PQuest is a III-V dedicated
etcher and I doubt the III-V users would be hot on you trying to develop a high
temp Pt process in the PQuest.

We have 3 high plasma density etchers (Lam, STS1 and STS2) but they are all
dedicated to Si etching. Note that the plasma density is important since it
determines the ion bombardment flux (ion current density Ji) and sputter rate
increases linearly with Ji. We have 2 medium density etchers (P5000 and
PlasmaQuest).  The P5000 is a "clean" tool and we are not kean on putting Pt
into it. There is no fundamental problem with trying to develop your process in
the PQuest. The III-V users let non-III-V users into the tool as long as the
process does not affect the existing processes. They are mostly not concerned
with metal contamination. However, the PQuest does have a limited bias voltage
capability (150 v range).  This bias may be enought since Kwon's 1998 Ar + Cl2
paper claims a bias of only 125V.   Note sputter rate increases with the sq root
of ion energy, which is determined by bias voltage.

The MRC and Dryteks are low density etchers. The MRC and Drytek4  have a RIE
configuration, which means they can develop high bias voltages. The Drytek 1 and
2 have a plasma mode configuration which means they are not good for sputter
etching.,

I f you want, we can meet and talk about using the PQuest for your needs.  Note
that the PQuest is a "dirty" tool and so you would be restricted in what tool
you can use following it use.

        Jim


Elizabeth Ann Hager-Barnard wrote:

> Hello,
>
>         My name is Lizzie Hager-Barnard and I'm a first year graduate
> student in Materials Sci. and Eng., working for Prof. Nick Melosh.
> (My SNF username is lizhb.)  I am trying to find an appropriate
> procedure for making trenches in my devices, which are layered
> silicon devices.  Specifically, my devices consist of SiO2, platinum,
> and SU-8 layers on silicon wafers.  (The trenches need to go through
> all of these materials.)  Since my devices contain platinum, I
> understand that they cannot be processed on any equipment classified
> as clean or semi-clean.  This poses a problem for me, as my advisor
> believes that a chlorine-based gas would be best for etching
> platinum, and all etching equipment using this type of gas is
> classified as clean or semi-clean, as far as I'm aware.
>         I spoke to Ms. Nancy Latta yesterday and she advised that I used the
> MRC etcher.  However she also suggested that I contact SpecMat to
> confirm the acceptability and effectiveness of this equipment for my
> device.  If you could advise me as to what etcher would be best in my
> situation, and let me know any steps that I need to take to make sure
> my procedure is acceptable, I would appreciate it.  If you have
> further questions about my produce, please feel free to contact me.
> I may be contacted at lizhb at stanford.edu or at (650)326-3844.
>
>         Thank you for your help,
>
>         Lizzie Hager-Barnard




More information about the specmat mailing list