Nowadays, high angle annular dark field (HAADF) scanning transmission electron microscopy (STEM) is one of the most popular materials’ characterisation techniques because of its ability to provide direct structural images at the atomic resolution [1]. Recently, Rosenauer et al. proposed a new imaging technique called imaging STEM (ISTEM) combining the conventional TEM imaging with STEM illumination [2]. This new, spatially incoherent imaging mode is particularly interesting as it provides direct structural images and visualisation of light elements, while it is robust towards chromatic aberrations. In this work, the ISTEM and STEM precision with which atomic column positions can be measured in a quantitative manner are compared.

In HAADF STEM imaging, statistical parameter estimation theory is an excellent tool to quantitatively extract structure parameters [3]. Here, an empirical model is fitted to an experimental image by optimising a criterion of goodness of fit. In this model, the shape of an atomic column is described by a Gaussian peak. The parameters of this empirical model can be linked to the unknown structure parameters of the material under study. The precision with which those parameters can be measured is mainly determined by the presence of shot noise and scan noise in the images. While post-processing techniques reduce both effects [4,5], scan noise errors and probe instabilities have no influence in the ISTEM imaging mode as images for all probe positions are integrated. Therefore, it is expected that without the use of post-processing techniques, atomic column positions can be measured more precisely for ISTEM imaging as compared to STEM imaging.

This assumption is tested on experimental images of PbTiO₃. The ISTEM image in Fig. 1a clearly resolves the light oxygen columns, while in the HAADF STEM image, Fig. 2a, these columns cannot be resolved. In order to extract the atomic column positions, Gaussian models are fitted to both images, shown in Figs. 1b and 2b. The precision with which individual column positions can be estimated is determined by calculating the standard deviation on the distance between the columns (Fig. 3). These results confirm that atomic column positions are measured more precise for ISTEM as compared to ADF STEM imaging. Furthermore, it demonstrates that the position of individual light atomic columns, like oxygen columns in the present example, can be estimated with a precision in the picometer range.

In conclusion, it is shown that ISTEM is a very promising technique for precise measurements of atomic column positions.

References

[1] Pennycook and Jesson, Physical Review Letters 64 (1990), p. 938

[2] Rosenauer et al., Physical Review Letters 113 (2014), 096101

[3] Van Aert et al., Ultramicroscopy 109 (2009), p. 1236

[4] Sang and LeBeau, Ultramicroscopy 138 (2014), p. 28

[5] Jones et al., Advanced Structural and Chemical Imaging 1 (2015), 8

The authors acknowledge financial support from the Research Foundation Flanders (FWO,Belgium) through project fundings (G.0374.13N, G.0369.15N and G.0368.15N) and a PhD grant to K.H.W. van den Bos. The research leading to these results has also received funding from the European Union Seventh Framework Programme [FP7/2007- 2013] under Grant agreement no. 312483 (ESTEEM2).

Figures:

Figure 1. (a) ISTEM image of PbTiO3 and (b) the refined Gaussian model with the estimated column positions for the different elements shown as an overlay.

Figure 2. (a) HAADF STEM image of PbTiO3 and (b) the refined Gaussian model with the estimated column positions for the different elements shown as an overlay.

Figure 3. Experimental precision with which the individual columns can be located from an iSTEM and HAADF STEM image.

« Back to The 16th European Microscopy Congress 2016

EMC Abstracts - https://emc-proceedings.com/abstract/precise-atomic-column-position-measurements-using-istem/