Defect free film deposition is one of the most challenging steps in CMOS manufacturing for 28nm process node and beyond. For precise control over surface, excellent uniformity and accurate material characteristics are required. IC manufacturers find it critical to monitor the film deposition process in terms of layer uniformity and other signatures/defects induced on the surface during the deposition process. In this work, an oxide layer deposited using flowable chemical vapor deposition (FCVD) process on a 300mm bare silicon wafer has been studied. Different process conditions with varying amount of precursor and reactant gases have been employed for the deposition of the oxide. The so called "haze" signatures and statistics found during inspection on the post-deposition surfaces are used to measure variation in surface quality, serving as a means to qualify the FCVD process.For this study, an FCVD tool fitted with two chambers with two sides each has been employed for the deposition of oxide. The integrated SURFmonitor feature of KLA-Tencor's SP3 -a blanket wafer defect inspection tool set -was successfully utilized to highlight the variation of the CVD tool's performance. Firstly, the overall study clearly demonstrates the capability of SURFmonitor to differentiate the process conditions between Process -1 and Process -2. Also, for a given process, it showcases the capability of differentiating between different deposition chambers. In addition, for a given process and specific chamber, grid analysis feature of SURFmonitor has been applied to highlight the within-spec and the out-of-spec areas for each of the wafers. The methodology described in this paper has already proven to be quite effective in monitoring the CVD process.
As dimensions shrink with each successive semiconductor technology node, the critical size of defects that can impact yield also shrinks. Monitoring the health of process equipment rigorously and regularly for key sources of contamination must be performed. Comparison of the critical feature size per technology node with the typical defectivity size used for process control shows a widening disparity over time. The need to close this gap is more important than ever in order to improve chip yields as the industry drives toward the 10nm and 7nm nodes. The latest generation defectivity inspection equipment can measure defect sizes close to technology critical feature size. Getting accurate data of real added defects is complicated for processes that involve film deposition due to decoration effects caused by the film conforming over the pre-existing defects. In this study, we bring out a clear methodology in optimizing the pre/post monitoring sizes of defects in order to minimize the effect of false adders. The relationship between pre and post sizing is well understood for mechanical checks. However, the impact on the sizing of defects because of film interaction is of critical interest. In this work, a series of different films and deposition methods are studied. Decoration effects of defects (as shown in Figure 1) are analyzed in depth and an extensive methodology of optimization of post-scan thresholds is elaborated. This methodology has proven to be extremely effective and has successfully been implemented at Global Foundries Fab 8 and 10. Figure 1. Impact on post-scan threshold because of decoration effect 206 978-1-5090-0270-2/16/$31.00
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.