Platinum has relatively melting and boiling points, so it must be heated
hotter than many other common metals for the same vapor pressure and hence
deposition rate. Because of this, at one place I worked, we evaporated just
100 A at a time, at 0.5-1.5 A/s. After each 100 A, we let the system cool
for about 30 minutes. I would guess that with your water-cooled substrate
holder, the holder might be running cooler, but there is probably a poor
thermal contact between the wafer and the holder.

For cleanliness, we always pumped to below 1e-6 Torr before the deposition.
We followed the standard practice of doing some evaporation with the
"shutter" over the sample to collect evaporated impurities before opening it
and exposing the substrate.

    --Kirt Williams

-----Original Message-----
From: mems-talk-bounces+kirt_williams=ieee.org@memsnet.org
[mailto:mems-talk-bounces+kirt_williams=ieee.org@memsnet.org] On Behalf Of
Andrew Sarangan
Sent: Thursday, March 31, 2011 2:08 PM
To: General MEMS discussion
Subject: [mems-talk] Platinum e-beam evap liftoff

I've been having trouble e-beam evaporating Pt on photoresist for
lift-off, at 0.5A/s. The resist seemed to be cracking and/or reflowing
after a few minutes, but it was ok since we only needed about 200A of
Pt. The substrate temperature also rises pretty quickly, to about
150C. Assuming the resist damage was due to excessive heat damage, I
installed a water-cooled substrate holder. Strangely, the resist
damage is significantly worse now. It takes just 10 seconds before the
resist becomes completely deformed, which can be visually seen through
the viewport. Except for the substrate temperature, which is at 30C,
everything else is identical during evaporation. Anyone has clues on
what this could be due to?