Comparison of SEM and HRTEM CD Measurements Extracted From Test Structures Having Feature Linewidths From 40 to 240 nm

Abstract

<para xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> Critical dimension (CD) measurements have been extracted from SEM and high-resolution transmission electron microscopy (HRTEM) images of the same set of monocrystalline silicon features having linewidths between 40 and 240 nm. The silicon features are incorporated into a new test structure that has been designed to facilitate this type of CD metrology study. Major improvements to previously reported HRTEM sample-preparation and fringe-counting procedures have been implemented. The purpose of this paper is to make a preliminary assessment of the calibration statistics of SEM transfer metrology when HRTEM is used as the primary metrology in CD reference material calibration. The linearity and the correlation of the regression between HRTEM and SEM measurements were very encouraging. However, further study of the calibration statistics, from which uncertainty estimates of the SEM CD measurements were obtained, revealed small but significant test-chip-to-test-chip variability of the SEM-to-HRTEM offset at the low single-digit nanometer level. Further measurements made the case that this unanticipated variability originated in the differences in the amounts of hydrocarbon deposition that were made by the SEM tool during the measurement cycle. This is considered to be a very useful finding because modern SEM tools, which can reduce hydrocarbon deposition below levels that were encountered here by almost an order of magnitude, are now becoming available. The results reported here provide a strong indication that HRTEM–SEM-based calibration approaches offer great promise for single-digit nanometer uncertainty. </para>