Guys, the apparent non-linear relation between layer thickness and breakdown strength is simply a statistical artifact related to material defect density. If you had a very thick, absolutely atomically perfect crystal layer (good luck with that), then the empirical breakdown difference measured between thick and thin layers would vanish.
Sydd, thanks for the links lately, very interesting :)
Programmer, Applied Science Technologist (AScT). Automated Test Engineering (ATE), shop 153, CFB Esquimalt, Box 17000 Station Forces, Victoria, BC, Canada V9A 7N2. My direct line: (250) 363-2061. Home: (250) 382-0068