Jianhua, I found that cake icing works very well for filling trenches. You dilute it with a little water and use a needle to inject it. Or pour it on and use a squeegee to remove the excess. Bake at 325F for about 10-15 minutes and it's solid (otherwise it will stay gel-like for a very long time - days). After patterning your PDMS (haven't done that myself), you can either bake the icing off at high temperature or soak the whole wafer in hot water for several hours. Best part, you can eat the rest of the icing if you get hungry. Or you can use a negative photoresist with the mask you used to define the trenches in the first place. Slow spin to reduce the effects of the trenches on uniformity. Might need a little hard bake after exposure and developing to prevent the solvents in the PDMS from messing up the photoresist. Acetone or hot NMP to remove the photoresist after patterning PDMS. Good luck. Mike Dr. Michael H. Beggans Naval Surface Warfare Center, Indian Head 101 Strauss Ave. Bldg. 302 Phone: (301) 744-1927 Indian Head, MD 20640 Fax: (301) 744-6406 -----Original Message----- From: Jianhua Tong [mailto:tongjh@student.dlut.edu.cn] Sent: Tuesday, July 25, 2006 11:34 PM To: mems-talk@memsnet.org Cc: jh.tong@utoronto.ca Subject: [mems-talk] Trench Filling and PDMS suspending Hi, all, I am designing a structure (a PDMS film bridge over a trench). I need to make a trench and fill it with some material(PSG,SOG,SU-8???...), and then spin PDMS on it and pattern. the trench is 1000um*1000um and 30-40 microns in depth. Can anyone tell me which material is fine for filling a 40um-deep trench, and what are the steps. I also want to know whether there are some good ways to etch PDMS. does O2 plasma work? I found a paper that used scratching and lift-off, but i dont think the surface will be very smooth after scratching. I just searched on the internet/IEL, but failed to find good ways on it. Thanks a lot, Jianhua ---------------------------- jh.tong@utoronto.ca tongjh@student.dlut.edu.cn