Industrial Challenges of TiN Hard Mask Wet Removal Process for 14nm Technology Node

Akihisa Iwasaki,Kristell Courouble,Steven Lippy,Fabrice Buisine,Hidekazu Ishikawa,Emanuel Cooper,Evelyn Kennedy,Stephane Zoll,Lucile Broussous
DOI: https://doi.org/10.4028/www.scientific.net/ssp.219.213
2014-09-01
Solid State Phenomena
Abstract:TiN Hard Mask (TiN-HM) integration scheme has been widely used for BEOL patterning in order to avoid ultra low-k (ULK) damage during plasma-ash process [1]. As the technology node advances, new integration schemes have to be used for the patterning of features below 80 nm pitch with 193 nm immersion lithography. In particular, thicker TiN-HM is necessary in order to ensure Self-Aligned-Via (SAV) integration which resolves via-metal short yield and TDDB issues caused by Litho-Etch-Litho-Etch (LELE) misalignment [2, 3]. The Cu filling process is significantly more difficult if the thick TiN is not removed because of the high aspect ratio of the structures. Moreover, with the use of TiN hard mask, a time-dependent crystal growth (TiCOF) residue may forms between line etch and metal deposition [4, 5], also hindering copper filling. Post-Etch-Treatment after line etching is one solution to the problem but N 2 plasma is not efficient enough to suppress the residue completely [6], and the CH 4 treatment proposed in [5] may be difficult to implement for 14 nm node, thus an efficient wet strip and clean provides a better solution.
English Else
What problem does this paper attempt to address?