Line edge roughness (LER) is a potential showstopper for the semiconductor industry. As the width of patterned line structures decreases, LER is becoming a non-negligible contributor to resist critical dimension (CD) variation. The International Technology Roadmap for Semiconductors (ITRS) specifies an LER control of 2.4 nm for 2005 and 1.3 nm for 2010. This level of control and the three dimensional nature of LER, pose a challenge to current measurement tools. LER control is a relatively new issue and there are questions on the best way to measure it, on whether the current tools can capture its real nature, and more importantly if we need to capture its real nature. The use of different roughness evaluation strategies and the possibility that different measurement conditions with the same instrument may give different values are also causes of concern. As a part of a study to determine an optimum way to evaluate LER, we measured a series of LER resist test structures made from a single level 193 nm alternating phase-shift reticle containing structures in clear field and dark field tones. The structure includes different periodicity, amplitudes, base linewidths, and sizes of assist features among others. Various types of AFM configurations are being used in order to determine optimum measurement conditions. This experiment requires repeated measurements using different sensor types, tips, scanning configuration and other AFM variables. Here, we present results from our measurements using two different types of tips on a top down scanned tip AFM, and our analysis using scale-space techniques .
Proceedings Title: American Society for Precision Engineering, Annual Meeting | 18th | Proceedings of ASPE 18th Annual Meeting | American Society for Precision Engineering
Conference Dates: October 26-31, 2003
Conference Title: ASPE--American Society for Precision Engineering
Pub Type: Conferences
atomic force microscope, line edge roughness, scale-space