Development and verification of an efficient simulation tool for electrically large doubly periodic 3D nanostructures in layered medium for optical Mueller matrix Fourier scatterometry (ScatteroSim)

• Applicant: Professor Dr.-Ing. Stephan Reichelt
• Subject Area: Measurement Systems
• Term: since 2025
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 553623432

Semiconductor chips remain at the core of modern science and technology, incessantly driving global competition. Device architectures such as FET structures, NAND flash memory, and DRAM are transitioning to three-dimensional (3D) designs. Optical scatterometry remains one of the most advanced methods for inspecting periodic nano-structures due to its precision, speed, and cost-effectiveness. However, retrieving structural parameters involves solving an inverse problem by comparing calculated and measured data. Without a robust simulation tool, this inverse problem cannot be solved, which has hindered the development of 3D optical scatterometry. The goal of this project is to develop and validate an efficient simulation tool for electrically large doubly periodic 3D nanostructures in layered media for optical Mueller matrix Fourier scatterometry. This tool is referred to as "ScatteroSim" henceforth. One of our main goals in this project is to simultaneously integrate periodic boundary conditions (PBC), the multilevel fast multipole method (MLFMM), and the periodic layered medium Green's function/thin dielectric sheet (PLMGF/TDS) approaches into a single simulation tool (ScatteroSim), based on the surface integral equation (SIE) method. Specific objectives include: (1): Developing a fast and rigorous simulation tool ScatteroSim for electrically large doubly periodic three-dimensional structures with high aspect ratio (lateral size to height ratio greater than 10) in a layered medium by implementing recently developed algorithms. (2): Applying ScatteroSim as a modeling technique to our recently developed white light Mueller Matrix Fourier Scatterometry (MMFS) setup to measure the structures mentioned in objective (1), which will also be fabricated in this project. Data sets will be generated, and a deep learning (DL) neural network will be trained. We are convinced that the developed software ScatteroSim will advance the field of 3D optical scatterometry, currently limited by the lack of a powerful modeling tool. Furthermore, it will greatly benefit scientific research on complex nanostructures such as photonic crystals, metamaterials, solar cells, and various other structures in the optical regime. It will also enable the simulation of periodic structures with surface roughness.

• DFG Programme: Research Grants
• Co-Investigator: Dr.-Ing. Karsten Frenner