Hello Peng Yao, The first thing is, yes, the wafer was probably dirty. However, PECVD nitride is not good for a KOH mask, anyways. It typically leaves pinholes, which will give the KOH avenues for attack during your bulk etch. Second, there are a couple of ways to wet etch Nitride. One is to dip it in concentrated HF for a long time (several hours, I think), which may cause your BOx layer some problems, another way is to use some variety of heated phosphoric acid. Both methods should be in the literature. Hope this helps, Jesse Fowler UCLA/MAE Dept., 420 Westwood Plaza, Room 37-129, ENGR IV Los Angeles, CA 90095-1597 | (310)825-3977 "Battery is safe if not provoked." -- _Batteries in a Portable World_ On Sat, 20 Apr 2002, Peng Yao wrote: > Hi, > I am trying to wet etch SOI wafer from back side and stop the etching at > the SiO2 layer. To do that, I first PECVD a 5000 A silicon nitride layer > on both side. Then I opened window on backside and used KOH to etch it. > What suprised me was that after one hour etching, the silicon layer under > the top silicon nitride protecting layer was attacked. And I can see a lot > of small crack formed in the silicon nitride layer, while the silicon > nitride layer on back side looked much better (But also have some crack). > There are two possible reason I am suspecting. One is that before silicon > nitride PECVD, I did not clean the wafer specially (take it out directly > from the shipping box and put it into the reactor). The other is, just my > guess, the difference thermal expand coefficient between top silicon layer > and the sio2 layer under it might form some small crack on the silicon > during the high temperature of PECVD process. > I don't know which one is more possible. And if it is the first one, does > any of you know how to get rid of the si3n4 layer which has been grown > without attacking the silicon layer underneath. > > Thanks a lot!!! > > Peng Yao > DOEs lab > Electrical Engineering Dept. > Univeristy of delaware > Newark D.E 19716 > _______________________________________________ > mems-talk@memsnet.org mailing list: to unsubscribe or change your list > options, visit http://mail.mems-exchange.org/mailman/listinfo/mems-talk > Hosted by the MEMS Exchange, providers of MEMS processing services. > Visit us at http://www.mems-exchange.org/