Characterization and elimination of threshold voltage problems in the Bimos process

Abstract

The author identifies two root causes and a number of special causes of high-P/low-N threshold voltage in the Bimos process. The first root cause was chemical carryover from wafer carriers exposed to HF. the F/sup -/ retained in the carrier can hydrate, form a vapor, and condense on the wafer. This phenomenon was the source of a large amount of unexplained high-P threshold problems in the past. The second root cause was oxide charge being introduced at the trim anneal furnace operation. Increasing the N/sub 2/ flow brought a major yield loss problem under control. since the two root causes have been eliminated, there have been no significant threshold voltage problems with this technology.<>