New Approach to Quantitative ADF STEM

Abstract

High-angle annular dark-field scanning transmission electron microscopy (HAADF STEM or Z-contrast) has been shown to be remarkably sensitive to atomic number (Z). However, HAADF images are currently formed on an arbitrary intensity scale, thereby limiting the possibility of truly quantitative imaging. Recently, it was reported that a mismatch exists between experimental and simulated image contrast in HAADF STEM [1]. Without an absolute scale, it is impossible to determine the cause of the discrepancy [2]. Additionally, an absolute scale would facilitate composition mapping at atomic resolution. Here we demonstrate that the HAADF detector can measure the incident beam intensity to normalize Z-contrast images onto an absolute intensity scale. We report on a practical approach that ensures that the detector does not saturate and is sufficiently linear over the intensity range of interest. An FEI Titan 80-300 STEM/TEM equipped with a super-twin lens (C s ~ 1.2 mm) operating at 300 kV was used for this study. We discuss the practical aspects of achieving quantified HAADF images. Characterization of the detector and acquisition/quantification of the image intensities will be described. In addition, the detection efficiency across the ADF detector has been investigated. As shown in Figure 1, the efficiency is not uniform and can vary drastically near the hole in the detector. By normalizing the atomically resolved signal, we demonstrate quantified HAADF imaging of a SrTiO 3 single crystal. Figure 2 displays an experimental image from a region ~ 200 Å thick and a corresponding Bloch wave image simulation that takes into account spatial incoherence. As can be seen, a quantitative match exists between simulations and experiments. Local sample thickness was determined by the electron energy loss spectroscopy (EELS) log-ratio method [3]. We will show that the combination of information from the HAADF background signal and the thickness determination by EELS can be used to provide improved estimates of thickness. 1. D. O. Klenov and S. Stemmer, Ultramicroscopy 106, 889 (2006).