A statistical model of signal–noise in scanning electron microscopy

A statistical model describing signal-noise generation and development along the signal formation process in a standard scanning electron microscope (SEM) using an Everhart–Thornley secondary electron detector is derived. Noise in the detector signal is modeled to originate from a cascade of five signal conversion stages. Based on the derived model, general conclusions are drawn concerning the total signal-to-noise ratio (SNR) at each stage, and the influence of each stage on the total SNR of the detector signal. The model is furthermore applied to a real-world SEM, and verified by experimental data. SCANNING 34: 137–144, 2012. © 2011 Wiley Periodicals, Inc.