Sorry for the delay in this response.  Hope it is helpful:

On Sunday, May 17, 1998 11:29 PM, Jonathan d trumbull
[SMTP:[email protected]] wrote:
> Regina
>
> > I got a "problem"- in the process I use for masking and dryetching my
> > photoresist (Az 5214) gets so "hardened that I can remove it with
> > conventional means such as aceton, developer or KOH.
>
> When it is hardbaked really "well" 5214 can be difficult to remove.  Did
> you try heating the acetone (~90C) and using ultrasonic agitation? If
> this doesn't work, NMP is also an option (I warm it to ~70C).  Also, if
> it won't hurt the underlying structure, 1:4 H2O2 and H2SO4 will pull it
> off.  5214 will go on pretty thin and that might make removal easier.
>
>  Jonathan Trumbull

        Troublesome resist, especially after ion implantation or RIE, will
generally have a hardened "skin" on the top surface.  This skin is believed
to be several hundred nm thick (depending on circumstances), and generally
is the part that makes resist removal so difficult.  The resist underneath
the skin is usually much easier to remove.  One can view the top surface
skin, as being a cross-linked polymerized layer that results from exposure
to implantation or plasma etching.  The material under the skin is generally
only modified by thermal affects, and hence may be removed with standard
methods.
        For resist like this, one can generally start with a short (1-5 min) RIE
with O2.  If one has access to an RIE with good back side cooling (Helium
cooling, for example), then use that to avoid further heating of the resist.
The etch time is intentionally chosen so that only 10-20% of the resist will
be etched.  This is done to break up and etch the skin, without removing the
bulk of the resist.  If the components of the top surface skin can be
removed by the O2 RIE, then the resist will be removed easily.  However, in
many resists there are inorganic components which cannot be volatilized in
an O2 plasma.  Hence, one tries to break them up and rinse them away, as
best as possible.
        After etching the top skin, a solvent-based approach is used, usually
with
acetone/isopropanol or NMP (a.k.a., N-Methyl Pyrolidone, or Shipley 1165).
If NMP is used, it is generally used at room temperature (for convenience
only), it will work much better if both heated and stirred.  After that a
final O2 etch is generally used, usually 5 min in a barrel asher (such as a
Branson) at 1000W (or whatever the standard resist strip process is).  The
final ashing may not be necessary depending on the application.
        Note that this procedure will work most of the time.  For resist that is
particularly bad (such as with resist that has been ion milled) the resist
sidewalls can also form a rather intractable skin.  These sidewalls can
leave "fences" of burned resist at the edge of all the resist patterns.  In
this case, one can still use the O2 RIE, and the solvent strip, however, we
will follow it with a heated wet chemical strip in Nanostrip (a stabilized
liquid piranha).  Follow the manufacturers instructions for heating it, and
use a stir plate to keep the liquid moving.  Heated Nanostrip works very
well, though it will attack aluminum aggressively -- Nanostrip at room
temperature attacks aluminum slowly.
        Regardless of the common assumptions, stripping resist can be non-
trivial
and can be very time consuming.

                        Good luck,

                        Kevin Shaw