Does anyone have experience using various dilutions of the AZ9260 resist.

I have been using the resist undiluted and in an in-house  4:1 and a 1:1
dilution with EBR, achieving FTs of 6im, 3im and 1im respectively.
With the 1im film I am trying to achieve a minimum feature (trenches) of 1im.
However the resist seems to be completely unstable. The process window for
exposure is 1-2 seconds and even this still varies along the typical 2x2cm2 area
of my usual sample size.
I understand that the AZ9260 has an aspect ration of 6:1, however I have been
unable to reliably resolve 1im features with it or any of it' s dilutions.

Another effect I am experiencing is that development rates for small features
seem to decrease around the optimum exposure time. That is to say, an
underexposed or overexposed sample requires longer than the optimised
development time.
Is this a behaviour to be expected of this resist?

A final issue with the resist is that both dilutions seem to need the same
exposure time. I haven' t managed to find the swing (interference) curves for
AZ9260 and was wondering if anyone knows where I could find this data.

An example of myr current process parameters (for FT: 1im) is as follows:
Mask aligner: Karl Suss MA4 with UV300 optics operating at 6mW/cm2
Spin parameters: 4000rpm, 60sec
Prebake: 100oC hotplate 1min/im
Re-hydration: 10min
Exposure: 53sec
Development: 1min/im

Has anyone experienced similar problems? Any useful suggestions?

Thanks!
Angie