论文部分内容阅读
光学发射光谱(OES)诊断技术是高密度等离子体刻蚀工艺过程的关键技术,OES信号作为一种实时信号可以用来预测刻蚀过程的进展程度和判断刻蚀性能的好坏。在自行研发的等离子体刻蚀机平台上,采用美国海洋公司的OES传感器系统,采集多晶硅刻蚀工艺中所产生的OES信号,利用主元素分析法(PCA)对OES数据进行压缩处理,提高了实时信号的快速处理能力。对实验数据的分析表明:波长为405nm的OES谱线可以明确显示出等离子体刻蚀进程,是一条表征等离子体刻蚀过程的状态检测及终点检测控制的特征谱线。在此基础之上,提出了基于PCA法的终点检测算法,用以判断刻蚀终点。
Optical emission spectroscopy (OES) is a key technology in high-density plasma etching process. OES signal can be used as a real-time signal to predict the progress of etching process and judge the etching performance. On the self-developed plasma etching machine platform, OES sensor system of American Oceanic Company was used to collect the OES signal generated in the polycrystalline silicon etching process, and the OES data was compressed by the principle element analysis (PCA) Real-time signal processing capabilities. The analysis of the experimental data shows that the OES spectrum with a wavelength of 405 nm can clearly show the plasma etching process and is a characteristic line characterizing the state detection and end point detection control of the plasma etching process. On this basis, the end point detection algorithm based on PCA method is proposed to judge the end point of etching.