Practical Electron Microscopy and Database

In 4D STEM measurements, managing spatial drift between successive acquisitions, e.g. spectrum-image acquisitions [1], is critical to ensuring accurate data. During the acquisition of each spectrum image (SI), drift correction was actively applied, which eliminated the need for post-processing corrections on individual spectrum images. The primary cause of spatial drift between successive spectrum images was the manual re-centering of the micropillar after each incremental stage rotation.

To quantify and correct this drift, High-Angle Annular Dark Field (HAADF) images were co-acquired during the experiment. HAADF imaging is chosen because it is less affected by orientation-dependent diffraction contrast, making it more reliable for detecting and correcting spatial drift. The correction in the X-Y plane is achieved through manual alignment of successive tilt-plane images. This manual method provided superior results compared to automated alignment techniques that use cross-correlation.

After applying these corrections, a detailed visual inspection of the features in the HAADF images indicated that the spatial-drift artifacts had been reduced to within 1–3 pixels normally [1]. This level of precision in drift correction is essential for maintaining the spatial fidelity of the data, particularly in experiments where the sample undergoes multiple rotations or other stage movements.

[1] Konrad Jarausch, Paul Thomas, Donovan N. Leonard, Ray Twesten, Christopher R. Booth, Four-dimensional STEM-EELS: Enabling nano-scale chemical tomography, Ultramicroscopy 109 (2009) 326–337.