Sequential tilting 4D-STEM for improved momentum-resolved STEM field mapping
Christoph Flathmann, Ulrich Ross, Jürgen Belz, Andreas Beyer, Kerstin Volz, Michael Seibt, Tobias Meyer · 2025
Read Sequential tilting 4D-STEM for improved momentum-resolved STEM field mapping by Christoph Flathmann, Ulrich Ross, Jürgen Belz, Andreas Beyer, Kerstin Volz, Michael Seibt, Tobias Meyer (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2505-23533.txt.
Introduction
Christoph Flathmann and co-authors present sequential tilting as a way to improve momentum-resolved scanning transmission electron microscopy. Their 2025 paper deals with a phase-contrast method that maps lateral magnetic and electric fields from micrometers down to atoms.
Fields a few nanometers to a few hundred nanometers wide often set device function, yet they are far weaker than atomic fields. Small changes in diffraction conditions can shift the measured signal a lot. Precession electron diffraction is one established fix, with the beam precessed continuously. Their alternative tilts the beam in a sequence of steps. The aim is cleaner field maps across material junctions.
Catalog record
| CREATOR | Christoph Flathmann, Ulrich Ross, Jürgen Belz, Andreas Beyer, Kerstin Volz, Michael Seibt, Tobias Meyer |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 29 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2505-23533
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Sequential tilting 4D-STEM for improved momentum-resolved STEM field mapping Christoph Flathmann, Ulrich Ross, Jürgen Belz, Andreas Beyer, Kerstin Volz, Michael Seibt, Tobias Meyer Abstract Momentum-resolved scanning transmission electron microscopy (MRSTEM) is a powerful phase-contrast technique that can map lateral magnetic and electric fields ranging from the micrometer to the subatomic scale. Resolving fields ranging from a few nanometers to a few hundred nanometers, as well as across material junctions, is particularly important since these fields often determine the functional properties of devices. However, it is also challenging since they are orders of magnitude smaller than atomic electric fields. Thus, subtle changes in diffraction conditions lead to significant changes in the measured MRSTEM signal. One established approach to partially overcome this problem is precession electron diffraction, in which the incident electron beam is continuously precessed while precession-averaged diffraction patterns are acquired. Here, we present an alternative approach in which we sequentially tilt the incident electron beam and record a full diffraction pattern for each tilt and spatial position. This approach requires no hardware modification of the instrument and enables the use of arbitrary beam tilt patterns that can be optimized for specific applications. Furthermore, recording diffraction patterns for every beam tilt allows access to additional information. In this work, we use this information to create virtual large-angle convergent beam electron diffraction (vLACBED) patterns to assess MRSTEM data quality and improve field measurements by applying different data analysis methods beyond simple averaging. The presented data acquisition concept can readily be applied to other 4D-STEM applications. 1 Introduction Nanometer-sized electric fields across junctions play a crucial role in electronic devices such as transistors [ 1 ] , solar cells [ 2 ] or batteries [ 3 , 4 , 5 ] , as the associated built-in potentials determine their functional properties. New concepts that improve the performance of these devices require miniaturization and increasingly complex designs. Thus, accurate measurements of the potentials on the relevant length scale of a few to about 100 nm are necessary for technological progress. In recent years, phase-contrast transmission electron microscopy (TEM) techniques have demonstrated great potential for accomplishing this task. One such method with a long tradition of mapping potentials across junctions is off-axis electron holography [ 6 , 7 ] . It has been used to study silicon p-n junctions [ 8 , 9 , 10 , 11 ] , compound semiconductors [ 12 , 13 ] and quantum wells [ 14 , 15 ] . However, off-axis electron holography has several drawbacks, such as the requirement of an undistorted reference wave and a limited field of view. An alternative approach to mapping fields involves measuring the deflection of the transmitted beam in scanning transmission electron microscopy (STEM) mode. In this case, one measures the average lateral momentum transferred from the sample to the electron beam via the center of mass (COM) of the diffraction pattern [ 16 ] , which is related to the lateral gradient of the projected electrostatic potential convolved with the incident probe intensity [ 17 ] . There are two methods for measuring the average momentum transfer: differential phase contrast (DPC) STEM and momentum-resolved STEM (MRSTEM). DPC uses a segmented detector to approximate the COM, allowing for the rapid acquisition of data, but is rather sensitive to the alignment of the experimental setup. Nevertheless, DPC has been used to map fields across p-n junctions and polarization fields in quantum wells [ 18 , 19 , 20 , 21 , 22 ] . With the advent of faster pixelated detectors, MRSTEM has become more common. Here, a full diffraction pattern is acquired at each scan position, enabling more robust data acquisition and post-processing than DPC. MRSTEM has been used to measure fields across p-n homojunctions with and without an applied bias, as well as potential drops across heterointerfaces [ 23 , 24 , 25 , 26 ] . However, all the phase-contrast TEM methods have issues with quantitatively and reliably measuring extended fields. This is because fields in the nanometer range typically have absolute values below 0.1 V/nm, which is orders of magnitude smaller than the gradients of atomic electric potentials. Thus, even small changes in the often dynamical diffraction conditions across the field of view lead to significant changes in the diffraction intensity, resulting in large and hard-to-predict signals in phase contrast TEM. This is particularly pronounced in complex samples, such as heterostructures, where strain gradients, defects, thickness variations, different lattice constants or symmetries, and local misorientation can strongly distort field measurements making it challenging to achieve accurate measurements in real devices [ 27 , 28 ] . The common approach to mitigate this issue involves combining DPC or MRSTEM with precession [ 23 , 26 , 29 ] . In this method, the incident electron beam continuously precesses around the optical axis at a polar angle of typically around 10 mrad. The precession is hardware-synchronized with the scan, and the diffraction signal is averaged over all azimuthal angles. This incoherent average of diffraction patterns has been shown to suppress dynamical diffraction effects as compared to individual diffraction patterns [ 30 ] , thus significantly improving the results of DPC and MRSTEM [ 26 ] . A comparison of the strengths and weaknesses of these convergent beam methods compared to off-axis electron holography using a plane incident wave can be found in reference [ 31 ] with the conclusion that for a proper crystal orientation a plane wave is beneficial since convergent beams more likely include incident directions with strong dynamical effects. The approach presented here goes beyond standard precession electron diffraction and offers the possibility to reduce these dynamical diffraction effects: It sequentially records full diffraction patterns for each scan position and tilt of the incident electron beam. This provides 6-dimensional data sets, which offer more information than previous methods. It also gives us the ability to use arbitrary tilt patterns, which has been reported to be beneficial for MRSTEM [ 29 ] . The underlying principle of this sequential tilting MRSTEM is very similar to the tilt scan averaged DPC STEM presented in references [ 32 , 33 , 34 , 35 , 36 ] , except that we have access to the full 6D data. In order to experimentally verify our approache, we use sequential tilting MRSTEM on AlGaAs/GaAs heterojunctions, a well-studied model system. We use the additional information contained in the sequential tilted MRSTEM data to assess the data quality, post-process the data for different sample orientations to reduce dynamical diffraction effects, and confirm that this additional information is essential to improving the reliability and quantifiability of MRSTEM results. 2 Experimental Figure 1: Beam tilt and detilt: a) shows the beam path without tilt. b) shows the beam path with tilt. c) shows the beam path with tilt, detilt and tilt induced shift correction. The relevant shifts are highlighted in the images. 2.1 Calibration procedure To achieve precise beam tilts without hardware modifications, we implemented a custom-developed calibration and acquisition procedure. A two-step calibration process is necessary to enable the sequential tilting MRSTEM: first, calibrate the beam tilts; then, calibrate the tilt-induced beam shift. Figure 1 a) - c) shows a schematic of the beam paths for