Research On The Gradient-based EUV Source And Mask Optimization Method

As the further shrinking of critical dimension of the large scale integrated circuit(LSIC), the next generation of lithography system named extreme ultraviolet(EUV) lithography system attracts the attention from the researchers in related field. It will replace the deep ultraviolet(DUV) lithography system in all probability, and become the main lithography technology in LSIC manufacture at 22 nm technology node and beyond. To promote lithographic imaging performance, resolution enhancement techniques(RET) are adopted which includes: optical proximity correction(OPC), off axis illumination(OAI), phase shifting masks(PSM), source and mask optimization(SMO), and so on. Among these RETs, SMO optimizes the source and mask at the same time. Compared with OPC, SMO would achieve better imaging quality since it has higher optimizing degrees of freedom.As an efficient RET, the research of DUV SMO algorithm is complete while EUV SMO is the opposite. This paper develops parametric simultaneous SMO, parametric hybrid SMO, pixelated simultaneous SMO and pixelated hybrid SMO aiming at the lithography system of 22-16 nm technology node and beyond, by employing an imaging model developed by our research group and gradient-based optimization algorithms. These methods consider the optical proximity effect, flare effect, photoresist effect and shadowing effect. The simulation results prove that the SMO algorithm can efficiently improve the EUV lithography system’s imaging fidelity. The comparison between SMO imaging performance and OPC imaging performance is also studied.