John, We observed when working with double coating process to make hyperthick (>250 µm) films of SU-8 that crosslinking of the SU-8 can begin thermally. This appeared to be a bottom up phenomena and presented itself as a thin layer of SU-8 that would not develop away in the unexposed regions. I am left wondering a few things... ....is this a chemical effect of casting a new layer of SU-8 onto an exposed layer of SU-8? The more I think about this the more I believe that you are inducing some lateral cross-linking between the first exposed layer and the second unexposed layer during the long second soft bake. ....if the blurry line you observe is a resist foot. It could be a response to the long soft bake for the second coating layer. Perhaps more optimization of the second soft bake will help? ....is a shift between exposures occurring? You didn't indicate what effect a single exposure of 80+ seconds has on the SU-8 process. Good luck. Best Regards, Garrett Oakes EV Group invent * innovate * implement Regional Sales Manager North America - Direct: +1 (480) 305 2443, Main: +1 (480) 305 2400 Fax: +1 (480) 305 2401 Cell: +1 (480) 516 6724 E-Mail: G.Oakes@EVGroup.com, Web: www.EVGroup.com -----Original Message----- From: John Hilton [mailto:spudcrazy@gmail.com] Sent: Thursday, November 13, 2008 6:33 PM To: General MEMS discussion Subject: [mems-talk] Blurring of SU-8 Layers upon development Hi, I'm making multilayer SU-8 molds and I have a recurring problem where the bottom layer appears to be blurry or reflowed upon development. This bottom layer is a 50um layer of SU-8 2050 (spun at 3000 RPM, soft bake 1min 65C 6min 95C, expose for 45s at 4.2 W/cm2 thru a long pass UV filter, PEB 1min 65C 6min 95C). The top layer is a 225um layer of SU-8 2100 (spun at 1350rpm, soft baked for a few hours, exposed for 82s at 4.2 W/cm2 in roughly 20 second increments to eliminate resist heating, followed by PEB and then development. Basically around all of my structures on the bottom 50um layer there is a blurry line when viewed in the microscope, between 10 and 100 um wide. Ashing w/ 02 plasma can remove some of this effect, but I don't know how that will effect the overall height of the mold and it makes it harder to de-mold PDMS later. Has anyone else had this problem? Testing has yet to indicate one single parameter that is the cause of the effect, but exposure time and bake times definitely come into play somehow. Vastly changing exposure times results in delamination of the bottom layer, however, and renders the final mold unusable. JP Hilton Graduate Student Columbia University BioMEMS Laboratory