计算光刻研究及进展  

Research and Progress of Computational Lithography

作　　者：马旭[1] 张胜恩 潘毅华 张钧碧 余成臻 董立松[2,3] 韦亚一 Ma Xu;Zhang Sheng’en;Pan Yihua;Zhang Junbi;Yu Chengzhen;Dong Lisong;Wei Yayi(Key Laboratory of Photoelectronic Imaging Technology and System of Ministry of Education of China,School of Optics and Photonics,Beijing Institute of Technology,Beijing 100081,China;Integrated Circuit Advanced Process R&D Center,Institute of Microelectronics of Chinese Academy of Sciences,Beijing 100029,China;School of Microelectronics,University of Chinese Academy of Sciences,Beijing 100049,China)

机构地区：[1]北京理工大学光电学院,光电成像与系统教育部重点实验室,北京100081 [2]中国科学院微电子研究所先导工艺研发中心,北京100029 [3]中国科学院大学微电子学院,北京100049

出　　处：《激光与光电子学进展》2022年第9期112-160,共49页Laser & Optoelectronics Progress

基　　金：国家科技重大专项(2017ZX02315001-003);国家自然科学基金(61675021)。

摘　　要：光刻是将集成电路器件的结构图形从掩模转移到硅片或其他半导体基片表面上的工艺过程,是实现高端芯片量产的关键技术。在摩尔定律的推动下,光刻技术跨越了90~7 nm及以下的多个工艺节点,逐步逼近其分辨率的物理极限。同时,光刻系统的衍射受限特性,以及各类系统像差、误差和工艺偏差,都会严重影响光刻成像精度。此时,必须采用计算光刻技术来提高光刻成像分辨率和图形保真度。计算光刻是涉及光学、半导体技术、计算科学、图像与信号处理、材料科学、信息学等多个专业的交叉研究领域。它以光学成像和工艺建模为基础,采用数学方法对光刻成像过程进行全链路的仿真与优化,实现成像误差的高精度补偿,能够有效提升工艺窗口和芯片制造良率,降低光刻工艺的研发周期与成本,目前已成为高端芯片制程的核心环节之一。本文首先简单介绍了计算光刻的前身,即传统的分辨率增强技术,在此基础上介绍了计算光刻的基本原理、模型和算法。之后对光学邻近效应校正、光源优化和光源掩模联合优化三种常用的计算光刻技术进行了综述,总结了相关的研究进展、成果和应用。最后,阐述了计算光刻当前所面临的需求与挑战,并讨论了最新技术进展和未来发展方向。Lithography is the process that transfers the structure pattern of integrated circuit device from the mask to the wafer or the surface of other semiconductor substrate,and it is the key technology to implement the mass production of high-end chips.Driven by the Moore’s law,lithography technique has stepped over multiple process nodes from 90 nm to 7 nm and beyond,gradually approaching the physical limit of its resolution.Meanwhile,the lithography image precision is seriously influenced by the diffraction limit property of lithography system,and various system aberrations,errors,and process variations.In this case,the computational lithography techniques must be used to improve the lithography image resolution and fidelity.Computational lithography is a cross research field that involves multiple professional domains,including the optics,semiconductor technology,computing science,image and signal processing,materials science,information science and so on.It is based on the optical imaging and process models,and uses the mathematical methods to simulate and optimize the entire lithography imaging chain,thus realizes the high-precise compensation of the image errors,and is capable of effectively improving the process window and chip manufacturing yield,as well as reducing the research and development cycle and cost of lithography process.To date,it has become one of the core links of the high-end chip manufacturing process.This article first briefly introduces the predecessor of computational lithography,i.e.,the traditional resolution enhancement technique,based on which the basic principles,models,and algorithms of computational lithography are introduced.Subsequently,three commonly used computational lithography techniques,including the optical proximity correction,source optimization,and source mask optimization,are reviewed,and the relevant research progress,achievements,and applications are summarized.Finally,this article expounds the current demands and challenges faced by the computational lithography,and

关 键 词：计算光刻 分辨率增强技术 先进半导体制造工艺 光学光刻 计算光学 光电图像处理 

分 类 号：O436[机械工程—光学工程]