Year: 2014
Type: conference-paper
Abstract: Low k1 and high NA lithography of critical layers require advanced resolution enhancement techniques for masks on large designs such as scattering bars, complicated mask edge placement and segmentation, etc. When mask feature sizes are comparable to or less than the exposure wavelength, the thin mask model (Kirchhoff) has been proven inadequate. It has been well known that mask 3D effects lead to image contrast loss, feature-dependent best focus, dependence of diffraction efficiency on illumination, etc. Mask transmission obtained with a rigorous solver of Maxwell Equations could catch these effects but running time is too slow and impractical for large chips OPC. HHA (Hybrid Hopkins Abbe) and DDM (Domain Decomposition Method) offer a fast and sufficiently accurate mask 3D OPC model solution. In this paper, we set up mask 3D model using HHA and DDM method for 28nm clear-field critical layers. We compare 3D model experimental results with thin mask 2D Kirchhoff model. The results shows mask 3D OPC model calibration get better model fitting error than 2D OPC model, also the improvement of model calibration results on model verification structures and on real design structures was realized by the use of 3D HHA and DDM models. The comparison result also shows process window data calibration accuracy increased by the mask 3D model.
Source: ECS Transactions
Institution: Semiconductor Manufacturing International (China)
Language: English
Subfield: Electrical and Electronic Engineering
Field: Engineering
Domain: Physical Sciences
Open Access status: bronze
OpenAlex ID: W1987074364
