Automatic Measurement of Silicon Lattice Spacings in High-Resolution Transmission Electron Microscopy Images Through 2D Discrete Fourier Transform and Inverse Discrete Fourier Transform

Abstract Line width (i.e., critical dimension, CD) is a crucial parameter in integrated circuits. To accurately control the CD in manufacturing, a reasonable CD measurement algorithm is required. We develop an automatic and accurate method based on a two-dimensional discrete Fourier transform for measuring the lattice spacings from high-resolution transmission electron microscopy images. Through the two-dimensional inverse discrete Fourier transform of the central spot and a pair of symmetrical diffraction spots, an image containing only a set of lattice spacings is obtained. Then, the pixel span of the lattice spacing is calculated through the centre of gravity method. Finally, we estimate the standard CD value according to the half-intensity method. The silicon crystal lattice constant guarantees the accuracy and traceability of the CD value. Through experiments, we demonstrate the efficiency of the proposed method, which can be conveniently applied to accurately measure CDs in practical applications.