Abstract
Optical proximity correction (OPC) has become an indispensable step in integrated circuit manufacturing. It requires a huge amount of calculation to obtain a sufficiently accurate OPC model and implement pattern correction. In this paper, the authors proposed an edge-based OPC method built on a vector imaging model, where the analytical correlation between the cost function and movement of each edge segment is established by the chain rule. First, the mask pattern is segmented and downsampled to get the mask image in order to reduce the total data. Second, the aerial image, various parameters on each evaluating point, and the final cost value are obtained in proper sequence. In each part of the OPC process, the procedures of solution and derivation are both recorded. After obtaining the cost value, the chain rule is applied, by which the differential relation between the cost value and movement of each segment is built. According to this differential relation, the next movement of each segment is decided under a quasi-Newton method. All results obtained by the proposed method are compared with results from commercial software. The comparison shows that the proposed OPC method has good OPC accuracy in few iterations.
© 2024 Optica Publishing Group
Full Article | PDF ArticleMore Like This
Guodong Chen, Sikun Li, and Xiangzhao Wang
Opt. Express 29(11) 17440-17463 (2021)
Xu Ma, Zhiqiang Wang, Yanqiu Li, Gonzalo R. Arce, Lisong Dong, and Javier Garcia-Frias
Opt. Express 26(11) 14479-14498 (2018)
Le Ma, Libin Zhang, Jianfang He, Huwen Ding, and Yayi Wei
Appl. Opt. 63(2) 499-505 (2024)