Autofocusing Image System of CD-SEM

Abstract

In order to realize auto focus during CD-SEM imaging process, the structure of image and scanning model is explained. The principle and the performance of image sharpness evaluation function, such as pixel gray variance function, gradient function, gray entropy function and SMD function, are analyzed in this paper. The experiment shows that all listed methods can evaluate the image sharpness correctly. But gray variance function excels in speed and evaluation effect, thus proving to be the ideal image sharpness function for CD-SEM auto focus. Keywords-Image sharpness, Evaluation function, Auto focus I. INTRODUCTION During the production of integrated circuit, mechanical or human factors may result in defects which can not be repaired, and lead to the sharp decline in yield. Therefore, process detection equipment is the critical technique for integrated circuit yield control. At the present time, the most widely used detection techniques in the production of integrated circuit are optical imaging technology, laser scanning analysis technology, Scanning Electron Microscope (SEM), Atomic Force Microscope (AFM), etc. Since the magnification of SEM can reach 100,000 times, and its resolution is much better than optical microscope, SEM is one powerful facility for observing and studying the microcosm. The integrated circuit detecting device based on SEM is called Critical Dimension-SEM (CD- SEM). CD-SEM is capable of measuring the one and two dimensional nanometer structures, and of continuous and lasting operation on the production line. Therefore it plays the leading role in detection equipment supplies for the semiconductor producers in and beyond 32 nm technology node. During the detection process, CD-SEM first acquires corresponding IC design pattern from the database, then moves the stage to the expected position for detection swiftly and accurately, after which the SEM auto focus process will be started. When the clearest SEM image is acquired, a comparison between SEM image and the design pattern will be proceeded to check for defect or measure relevant parameters, such as Line-Edge Roughness, Line-Width Roughness, line width, contact hole, sidewall angle and curvature radius, etc. The products will be checked for qualification according to whether these parameters could meet the set standards or not