Stochastic Patterning Simulation Using Attenuated Phase-Shift Mask for Extreme Ultraviolet Lithography

Abstract

In extreme ultraviolet lithography, the photon shot noise effect is a main cause of low-quality imaging characteristics such as line edge roughness and critical dimension (CD) nonuniformity. In this study, the stochastic imaging property of an attenuated phase-shift mask (PSM) was evaluated, and the results showed that “informative” photons from the first order diffraction are essential for mitigating the photon shot noise effect. This structure exhibits a reflectivity of ∼6% at the absorber stack and a phase shift of 180° at 13.5 nm wavelength. The improved stochastic patterning properties of the PSM were compared with those of a conventional binary intensity mask.

Attenuated phase-shift mask for mitigation of photon shot noise effect in contact hole pattern for extreme ultraviolet lithography

“…This result is consistent with our previous reports for L/S patterns, where the PSM shows greater improvement for smaller patterns. 13) Similarly, improvement is also observed in the mask image contrast and ILS, as shown in Fig. 1(c).…”

“…19,20) This is speculated to be the main reason for the smaller dose-to-size results for the PSM. 13) The visible results of absorbed photon distribution in the photoresist for 25 contact holes and the magnified result for a single contact hole of a 22-nm hp C/H pattern are shown in Fig. 3.…”

“…Recently, we have demonstrated that only the photons that affect the LWR of L/S patterns are diffracted, and the attenuated PSM shows improved LWR due to a larger number of diffracted photons. 13) Since the diffracted photons provide information concerning pattern images, it is expected that improving the diffraction efficiency of the mask will be an effective way to alleviate the PSN effect in C/H patterns. Therefore, in this study, we investigate the effect of an attenuated phase-shift mask (PSM) on C/H patterning performance parameters, such as CDU and CER, through comparison with a binary intensity mask (BIM) using stochastic simulation of C/H patterns.…”