Unstructured tetrahedric meshes for the description of complex three-dimensional sample geometries in Monte Carlo simulation of scanning electron microscopy images for metrology applications

• Scanning electron microscopy image simulation.
• 3D Monte Carlo simulation of complex meshed structures.
• High-performance computing and parallelization.
• Optimization of the computing time.
• Secondary Electron image synthesis.

Monte Carlo simulation codes for the reconstruction of secondary electron images in scanning electron microscopy for metrology applications require a very accurate description of the sample geometry, which cannot be achieved with the traditional approach using basic bodies. In present paper, unstructured tetrahedric meshes are introduced for the first time to define three-dimensional sample geometries in the nanometre scale to account among other for complex shapes and roughness. An original hierarchical description is developed to keep the simulation time within a factor of two or three as obtained by basic bodies. Finally, the performance of the proposed approach is demonstrated quantitatively based on representative metrology benchmarks.