There are two other possibilities that I see. First of all you could try heating up the sample close to the glass transition temperture. We've done that and it seems to work. The sample would perform sort of a reflow then. The other possibility would be using a negative resist where you can perform grayscale lithography to get a round profile into your resist that is then transfered into your silicon by etching. All the best, Marcel Jaibir sharma schrieb: > Dear Chen, > > You can try dry thermal oxidation and then etching the oxide with the help of BHF. > I am not very sure with this but feel it may solve your problem. > > with regards > jaibir > Marcel Spurny, Dipl. Phys. MPhys. School of Physics & Astronomy University of St. Andrews North Haugh St. Andrews KY16 9SS Fife Scotland, United Kingdom Tel.: 01334 467336