Roughness Decomposition: An on-Wafer Methodology to Discriminate Mask, Metrology, and Shot Noise Contributions

摘要

In order to meet the tight Line Width Roughness (LWR) requirements for advanced metrology nodes, it is critical tobe able to identify what the fundamental sources of roughness are, so that they can be individually minimized. Infact, more and more efforts aiming to decouple mask and / or metrology contribution from wafer data have beenrecently reported . However, these approaches often rely heavily on extensive mask characterization, somethingthat is not always easily available.We propose here an alternative path to investigate and discriminate the root causes of LWR using only wafer data. Itis based on Local Critical Dimension Uniformity (LCDU) decomposition , a methodology used to identify andquantify the individual LCDU contributors. The decomposition approach requires a smart sampling of the waferprint, in which an array of contact hole is measured in different dies multiple times. For such an approach to besuccessful, it is critical to ensure that the measurement locations are individually identified. Hence, it is necessary toanchor the metrology to a reference feature. A linear nested model is then used to quantify the three mainvariability components (mask, shot noise, and metrology). This approach allows to sample thousands of features atmask, a task that would not be practically achievable through direct mask measurements.In this work, LWR decomposition is implemented for the first time. To this aim, 18nm lines at 36nm pitch, printedby EUV lithography, were used. We specifically worked with a pattern including programmed defects, used asanchoring features for the metrology. In order to limit the impact of the metrology noise, expected to be higher forlines as compared to CH, we sampled over 8000 anchored measurements per image (in the CH case, only 81measurements per image were needed). The LWR decomposition results indicated the dominance of the metrologynoise, as expected. In addition, the mask contribution was observed to be less relevant that the shot noise.To verify the accuracy of the LWR decomposition results, Power Spectral Density (PSD) analysis on wafer andmask SEM images was used. The metrology noise contribution was removed at both mask and wafer level using anun-biasing normalization of the PSD curves . The comparison with the PSD analysis confirmed the feasibility ofLWR decomposition, opening the way to a more effective diagnostic technique for roughness and stochastics.