As is well known, Mr. Moffat is an expert on image reversal (YES systems),
however I would add that I have found a single layer of image-reversed AZ5214E
can be problematic in photolithography for liftoff in sputtering systems (not
ebeam) and that may have larger targets, i.e., larger solid angle subtended at
sample by sputtered species (and also for the case when one needs sputtered
films thicker than the spun AZ thickness). I have developed a LOR/AZ5214E image-
reversed bilayer that is robust and reproducible for  sputtering with a large 8"
Al target (target-to-sample distance of approximately 10 inches (when I do Al
liftoff) and for up to 1 micron thick Nb (for which one cannot do e-beam), with
a 2" Nb target. In each case, the sputtered layers easily lift off in only a few
minutes soak in Remover PG. The downside is a double spin and three softbakes
and flood expose, but this recipe is bullet proof. I'll be giving more details
in a paper soon.

Thomas