Pondering again, I wonder whether the solvent has been completely driven
out of the PMMA during the bake?  If not, maybe there's a mechanism
which would then cause cracking at a later stage?  Just wondering.
Wouldn't explain the insolubility, of course.

Andy

On 03/06/2010 09:25, Andrew Irvine wrote:
> Hi Lando
>
> To add a little bit to the discussion, I usually bake my PMMA at 120 for
> 1-3 hours, but in the past have baked at 180 for 1 hour and have never
> observed cracking except for seriously thick resists (2-3um from
> memory). There's a bit of degradation of normal (sub-micron) films after
> 12-16 hours at 180C. Of course, one hotplate's 180C might be another
> oven's 200C..? However, I have not seen any noticeable difference in
> performance across those ranges.
>
> But that's a bit of a red herring if I'm following this correctly -
> Lando, your resist is not cracked after the plasma descum, and not
> cracked when you load into the e-beam evaporator, but cracked when you
> unload? If so, forget resist bake temperature/time. Cracking says it's
> been thoroughly roasted in the evaporator to me, and presumably (well)
> above 200C. Insolubility says the same (but could also be massive UV or
> e-beam exposure, though both seem implausible).
>
> I'm finding it hard to resist the temptation to type this in the style
> of Hercule Poirot for some reason. Anyway...
>
> You say "most of my colleagues are using photoresist" - do your
> colleagues do the exactly same process apart from a different resist? I
> don't have experience of all the photoresists in the world, but I'd
> generally expect PMMA to cope with higher temperatures than the average
> photoresist. Something doesn't quite add up. Are your colleagues doing
> *exactly* the same, or is there some subtle difference you can think of?
>
> As others have said, you can usually lift off a sputtered layer way more
> easily than you'd expect, particularly if it's just 20nm thick. I'd give
> that a go, and then we can think some more.
>
> But in summary, my bet (in agreement with some other contributors) is
> that it looks like you're cooking your sample during the e-beam
> evaporation, even if your temperature monitor tells you otherwise.
>
> Andy
>
> p.s. I'm still etching MnAs badly, but it's improving... ideas still
> welcomed!