Wave Front Sensor for Highly Accurate Characterization of Flatness on Wafer Surfaces

Semiconductor manufacturing processes start on bare silicon substrates having excellent surface flatness. In the subsequent process chain of manufacturing of integrated circuits, it is crucial to maintain this initial surface flatness. New lithography technologies, as for example immersion lithography and Extreme Ultra Violet (EUV), require high flatness of the exposure surfaces as e.g. the depth of focus is impacted. Integrated device manufacturers use different technologies involving materials such as metals, semiconductors, and isolators to build three-dimensional structures on the wafer. The miscellaneousness of materials on the surface of wafers with highly integrated circuits limits the use of metrology developed for flatness analysis of bare silicon wafers. Nevertheless, it is essential for semiconductor manufacturing to measure and control the topography of wafer surfaces at nanometer scale. Wave front sensing was developed to characterize topography on bare and patterned wafer surfaces. Post processing of the acquired data applied 2 D Gaussian high pass filters to obtain the flatness data. The standard deviation of the flatness data after filtering was almost independent with respect to the lateral resolution. On bare wafer surfaces, the standard deviation was found to be below 1 nm, on patterned surfaces to be below 10 nm. Both, changes in reflectivity and filtering were found to impact and limit the accuracy of the determination of flatness. The accuracy of flatness measurement was found to be better than 20 nm. The results were validated by using a wavefront sensor according to the method of Makyoh. The technique has a high potential to provide high speed, contactless, and local inspection of flatness on bare and patterned wafer surfaces