Uncertainties of model fits in EUV nanograting metrology

• Applicant: Dr.-Ing. Sascha Manuel Brose
• Subject Area: Measurement Systems ; Electronic Semiconductors, Components and Circuits, Integrated Systems, Sensor Technology, Theoretical Electrical Engineering
• Term: since 2022
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 511573344

The characterization of nanogratings that are embedded between the functional elements of an integrated circuit wafer is one major semiconductor metrology task addressed with techniques such as scatterometry. The nanogratings serve as metrology targets resembling important features of the surrounding functional elements while offering a higher information density through their periodicity. As feature sizes in semiconductor manufacturing continue to shrink and architectures become increasingly three-dimensional, model-based optical metrology techniques currently used for process control and quality assurance using IR, VIS or UV light as the probing radiation (e.g. scatterometry), approach the limits of their capabilities. A promising alternative probing radiation regime for this application is the extreme ultraviolet (EUV). In order to compare model-based nanograting metrology techniques, the uncertainty regarding the measurands (e.g. grating line width and grating line height) needs to be quantified. A number of uncertainty sources in the underlying physics, the experimental measurements as well as in the utilized model contribute to the final aggregated uncertainty of a measurand. Identifying, understanding and aggregating these main sources of uncertainty in EUV-based nanograting metrology is the main scope of this project. Firstly, correlations of the measurands are one source of uncertainty in a model-based metrology technique. Since correlations in EUV nanograting metrology are specific to the interaction of EUV radiation with the nanograting, correlations and their effect on the final measurand uncertainty are quantified as part of this project. Secondly, the implementation of roughness effects in the model is a possible source of a model error. Stochastical roughness effects are often hard or even impossible to account for in computational models. This is due to some necessary simplifications, e.g. periodic boundary conditions. Thirdly, the optical constants used in the model are a possible source of a model error. At visible wavelengths, optical constants of very small structures on the nanoscale (< 30 nm), as occurring in semiconductor manufacturing, have exhibited a dependence on the structure size. Examining such a behavior and quantifying it for EUV wavelengths is part of this project. A successful outcome of the project will yield a comprehensive understanding of the main contributors to the uncertainty in model-based EUV nanograting metrology. Specifically, a quantitative framework for aggregating the different uncertainty sources will be developed, applied to various industry-relevant application cases in metrology and validated through experiments. The results will serve as means to benchmark EUV nanograting metrology to other competing metrology techniques.

• DFG Programme: Research Grants