We currently use cyclized negative resist (TOK OMR-83) for trace
patterning.  We have been on some occasions experiencing residual negative
resist after xylene develop and n-butyl rinse.  The residue appears to have
bled off the cross-linked pattern (see
http://s277.photobucket.com/albums/kk74/microfiz/  for an image sample).
Our current process involves 1.0 um thick negative resist, 30 minutes of
softbake at 110C and 60 mJs of exposure (i-line).  The residue does appear
cross-linked because it cannot be removed by redevelop.

      We have several suspicions (namely insufficient expose, insufficient
develop, errors in coating, etc) but single-variable matrices have not
pointed to any strong root cause.  Has anyone had any experience with this
type of resist?  Thanks.