Scatterometry-based assessment of defective nanograss surfaces - a virtual experimentation setup for uncertainty analysis

Defects in nanosurfaces have to be detected in or close to their production process. For this purpose, scatterometry promises a non-contact approach with in-process capability. However, a reliable defect inspection of nanosurfaces with randomness such as the random positions of the rods of ZnO nanograss is challenging due to the inherent surface stochastics. To clarify the potential of a scatterometry-based classification of such defective surfaces, a statistical evaluation in combination with a virtual experimentation is proposed. The virtual experiments consist of a) multiple surfaces with the same defect grade, b) simulations of the surfaces’ scattered light distribution, and c) data processing for obtaining the virtual measurement results. The proposed approach is demonstrated for ZnO nanograss surfaces, where the vacancy of nanograss is the studied defect as an example. As a central result, the surface-randomness-based minimum achievable measurement uncertainty is 20 % of the maximum defect value. Furthermore, the proposed method is universally applicable for any type of nanostructure/defect, to clarify the classification potential of surfaces with scatterometric measurements. 54 d Charakterisierung von Nanostrukturen Ulrike Böhm D O N N E R S T A G