Monte Carlo Simulation Study of Scanning Electron Microscopy Images of Rough Surfaces

Y. G. Li,S. F. Mao,H. M. Li,S. M. Xiao,Z. J. Ding
DOI: https://doi.org/10.1063/1.2977745
IF: 2.877
2008-01-01
Journal of Applied Physics
Abstract:In this paper, we have developed a Monte Carlo (MC) simulation method for calculation of scanning electron microscopy (SEM) images of rough surfaces. The roughness structure is constructed in a finite element triangulated mesh by using a Gaussian function to describe the distribution of amplitude of the random rough peaks. Further spatial subdividing can accelerate the calculation and improves MC simulation efficiency. The MC model is based on the using of the Mott cross section for description of the electron elastic scattering and the using of the full Penn algorithm in a dielectric functional approach to the electron inelastic scattering. This simulation relates directly a defined rough surface structure modeling described by exact values of roughness parameters to the contrast observed in a SEM image, enabling the investigation of the influence of line edge roughness to the critical dimension (CD) metrology of a metal-oxide-semiconductor device by SEM. Example calculation of line images with sidewall roughness demonstrates that the present MC simulation method is useful for CD metrology of nanostructures by CD-SEM and, especially, for the linewidth measurement in the integrated circuit industry.
What problem does this paper attempt to address?