Effect of thermalization distance on chemical gradient of line-and-space patterns with 7 nm half-pitch in chemically amplified extreme ultraviolet resists

Abstract

In the development of high-resolution chemically amplified resists used for extreme-ultraviolet (EUV) lithography, the thermalization distance of secondary electrons is a significant concern. In this study, the effects of thermalization distance on the chemical gradient of line-and-space patterns with 7 nm half-pitch were investigated on the basis of the sensitization mechanisms of chemically amplified EUV resists. For the acid generator concentration of 10 wt % and the sensitivity of 30 mJ cm %2 , the normalized chemical gradient monotonically increased when the thermalization distance decreased from 6 to 1 nm. For the acid generator concentration of 30 wt % and the sensitivity of 65 mJ cm %2 , the normalized chemical gradient had a peak value at the thermalization distance of 1.5 nm. It was found that the optimum thermalization distance in terms of normalized chemical gradient depended on the acid generator concentration and sensitivity.

Effect of thermalization distance on stochastic phenomena in 7-nm-half-pitch line-and-space pattern fabrication using chemically amplified extreme ultraviolet resists

“…Note that "average" means average Deprotonation efficiency of proton source 44) 1.0 Deprotonation efficiency of nonproton source 34) 0.3 in the line direction in this study. As the dependence of the initial acid distribution on the thermalization distance has been reported, 25) those profiles are not shown here. The representative profiles of protected unit distributions calculated with thermalization distances of 5, 3, and 1 nm are shown in Fig.…”

“…24) We have investigated the dependence of acid image log slope and the gradient of protected unit concentration on thermalization distance, assuming 7-nm-half-pitch line-andspace patterns. 25) Although the effect of thermalization distance was clarified from the viewpoint of image quality, 25) the stochastic phenomena become a critical issue with the reduction in feature size. [26][27][28][29] It has been reported that the probability of defect generation more rapidly increases with a reduction in feature size than the value of LER.…”