Self-Aligned Double Patterning Decomposition for Overlay 自对准双图案分解叠加.pptVIP

Self-Aligned Double Patterning Decomposition for Overlay Minimization and Hot Spot Detection H. Zhang, Y. Du, M. D.F. Wong, R. Topaloglu Dept. of ECE, University of Illinois at Urbana-Champaign Outline Introduction Overview of 2D SADP Process Layout Decomposition Problem Formulation Problem Reduction Experimental Results Conclusions Introduction Double patterning lithography (DPL) is the enabling technology for printing in sub-32nm nodes DPL technologies can be classified into two major types: Double-exposure double-patterning (DEDP) Single-exposure double-patterning (SEDP) Self-aligned double patterning (SADP) Introduction In SADP process, given a target layout, before the manufacturing, core mask and trim mask must be generated from the intended target layout. Unlike the DEDP, the core mask and trim mask are not always directly from the original layout. The problem of generating the core and trim mask from a 2D designed layout is called SADP decomposition. Introduction Overview of 2D SADP Process Layout Decomposition Problem Formulation Feature Generation and ILP formulation Layout Decomposition Problem Formulation Feature and non-feature region constraint: Feature is true if and only if this location has trim mask and no side-wall. Non-feature region: Layout Decomposition Problem Formulation Core and trim mask geometry constraints: Sidewall adjacency rule: Si is 1 if and only if Ci is 0, among the core variables{Cj, Cj+1,…,Cj+m} within the sidewall width distance, there is at least one variable equal to 1. Layout Decomposition Problem Formulation Minimum corner-corner rule: If Ci is 1, Ci+1 is 0 and Ci+2 is 0, then all variables {Ck, Ck+1, …, Ck+q} within the minimum corner-corner distance should be 0. Layout Decomposition Problem Formulation Minimum space rule: If Ci is 1 and Ci+1 is 0, then all variables {Ci+2, Ci+3, …, Ci+p} within the minimum space distance should be 0. Layout Decomposition Problem Formulation Minimum width rule: If Ci is 0 and Ci+1 i