Energy‐filtered transmission electron microscopy of multilayers in semiconductors

Abstract

Quantitative analysis of novel semiconductors with wide or ultrathin multilayers of atomic dimensions is very important in order to control electronic and optical properties, but rather difficult due to the limited resolution in most techniques. In this paper we attempt to assess how effectively the total As dopant concentration in ultrathin As doped layers in InP and the Ti atomic fraction in a TixAl1-xN multilayer can be analysed quantitatively using energy-filtered imaging. These two materials have characteristic edges located at widely different energy losses, with the L edge of As being above 1000 eV, while that of Ti is around 450 eV. We have quantified the As concentration using the three-window technique and theoretical cross-sections and we find that the resolution limit is dominated by the signal-to-noise ratio in this delta-doped specimen. However, the accuracy of the Ti atomic fraction in TixAl1-xN can be as good as 10 at% for specimens of uniform thickness made by focused ion beam milling. We will compare our results with measurements of the composition made using Fresnel contrast, high resolution imaging and high angle annular dark field techniques.

A new algorithm to reduce noise in microscopy images implemented with a simple program in python

“…Even the primary electrons by transmission electron microscopy (TEM) show shot noise that means that the number of electrons hitting a detector during a given pixel time is statistically distributed on a more or less Gaussian basis (Reimer and Kohl, 2008). The resolution itself is controlled also by the signal-to-noise ratio (Liu et al, 1999).…”