极紫外光刻
计量学
临界尺寸
平版印刷术
薄脆饼
表征(材料科学)
计算机科学
采样(信号处理)
维数(图论)
集合(抽象数据类型)
可靠性工程
电子工程
材料科学
物理
纳米技术
光学
工程类
电信
数学
探测器
程序设计语言
纯数学
作者
Harm Dillen,Yi‐Hsin Chang,Fei Wang,Marc Kea,Gijsbert Rispens,Marleen Kooiman,S. Hunsche,Fu‐Ming Wang,Daniel Tien,Peng Tang,Pengcheng Zhang
摘要
Traditionally, the performance of a lithography or patterning step is described by its mean size and the spread at a 3 sigma probability. Recent papers by Bristol, Brunner and others have shown this is insufficient to describe the process capability in EUV lithography. To address this challenge, an enormous increase of sampling CD (critical dimension) values is needed to describe the actual distribution on the wafer. We will show how we can address this by leveraging the HMI eP5 e-Beam system to acquire a set of CDs of previously unknown size. We will further show that extended sampling leads to better understanding of this phenomena, as we can probe full distribution behavior even on a limited number of repeated exposures on a wafer.
科研通智能强力驱动
Strongly Powered by AbleSci AI