Direct Measurement of Sidewall Roughness on Si, Poly-Si and Poly-Sige by AFM

Valur Gudmundsson,Per-Erik Hellstrom,Shi-Li Zhang,Mikael Ostling
DOI: https://doi.org/10.1088/1742-6596/100/6/062021
2008-01-01
Journal of Physics Conference Series
Abstract:In this paper the effect of the commonly used HBr/Cl(2) chemistry for dry etching on the line-edge roughness (LER) of photoresist patterned single crystalline Si (sc-Si), polycrystalline Si (poly-Si) and poly-Si(0.2)Ge(0.8) sidewalls was characterized. Measurements were done by means of atomic force microscopy in combination with an elaborated sample preparation technique that allowed the LER at different depths of the sidewall to be measured. Samples were patterned by I-line lithography and etching was performed at an RF power of 200 W using HBr/Cl(2) (30/10 sccm) plasma. For sc-Si the photoresist and Si sidewalls had an LER of 0.8-1.4 nm and 1.5-2 nm, respectively. For poly-Si and poly-SiGe the photoresist sidewall roughness was, respectively, increased to 1.5-3 nm and 2-3.5 nm due to light scattering from the rough surface of the polycrystalline materials. The poly-Si film had a sidewall roughness of 3-4 nm. Poly-SiGe sidewall exhibited larger roughness with an LER of 5-12 nm which was not transferred from the photoresist. The results show that for sc-Si and poly-Si the sidewall roughness mainly originates from the photoresist process and little additional roughening is caused by the HBr/Cl(2) etching. However, for poly-Si(0.2)Ge(0.8) the LER is considerably increased from that of the photoresist indicating that the HBr/Cl(2) etching is the main contributor to the LER.
What problem does this paper attempt to address?