STEM imaging consists of scanning a tiny electron beam over a sample and capturing a fraction of the scattering signal using one or more integrating detectors. These detectors are mostly circular, annular, or – rarely – divided into a few segments.
4D STEM follows the STEM imaging approach, but it replaces the conventional integrating detectors with a pixelated one. With this, the entire scattering distribution – or diffraction pattern – is recorded at every scanning position, and valuable information can be recovered freely in the data analysis.
4D STEM at a fast speed can enable a number of advanced analyses, which are often not possible with conventional STEM detectors. How fast is this speed, you ask? As fast as over 100,000 diffraction patterns per second – and comparable to a typical STEM acquisition setup’s dwell time.
- Conduct 4D STEM acquisition at speeds that exceed sample drift and/or damage. Matching the well-established STEM standard with a 10-µs dwell time is now possible.
- Sensitivity and a high dynamic range allow for simultaneous handling of high intensities (the central spot) and low counts (high-angle scattering intensities). This is essential for recording clear diffraction patterns.
- The pixel size is optimized for greater sensitivity (a high DQE) and signal localization (MTF).
4D STEM Techniques
Virtual STEM Imaging
Imagine that you could flexibly change your STEM detector’s size and shape to cover different parts of the scattering distribution. And that you could do this after the experiment, without scanning the electron beam over the same sample region again.
With 4D STEM datasets, it is possible to synthesize STEM images by selecting your desired regions from the recorded diffraction patterns. The use of virtual masks to select regions of interest flexibly allows emulation of customized detector geometries, equivalent to the traditional STEM modes and beyond.
Electron scattering from a polycrystalline gold sample, and reconstructed virtual BF and ADF after a 4D STEM acquisition. (a) The sum of diffraction patterns contained in the 4D STEM acquisition. (b) and (c) Virtual BF and ADF images, respectively, using the indicated masks. Acquired with DECTRIS ARINA.
Differential Phase Contrast (DPC) for Magnetic and Electrostatic Fields Mapping
DPC has proven to be a great way to measure a sample’s electromagnetic fields. It aids the visualization of magnetic domains, or the mapping of local electrostatic fields down to the atomic columns.
Coupled with smart data analysis, DPC can be used to reveal light atomic species. And the flexibility provided by a 4D STEM acquisition allows for the most precise data analysis.
Comparing the visibility of light atoms in a SmB6 crystal sample. (a) Virtual BF with a SmB6 atomic structure overlay. (b) A DPC vector field with color coding to indicate direction and modulus. (c) A calculated iCoM image from the previous DPC. All three images were calculated from a single 4D STEM acquisition. Acquired with DECTRIS ARINA. Acknowledgment: Elisabeth Mueller (PSI) and Mingjian Wu (FAU).
This advanced imaging technique can uncover the most detailed view of the specimen under investigation. Ptychography is a computational image reconstruction technique that explores the interference between the direct electron beam and its diffracted components.
The most recent electron ptychography results allow for a posteriori correction of optical aberrations, permitting microscopists to achieve an unmatched spatial resolution.
Also relevant to engineering materials, particularly when mechanical and electronic properties are key, is the local strain at interfaces or structural defects.
The existing compression, or stretching, of interatomic spacing is also encoded in the diffraction patterns that are recorded in a 4D STEM measurement. Given its potential to map properties with nm-scale resolution, this approach is promising. It will support the strain-engineering approach that is currently used in the semiconductor and optoelectronic industries – just more straightforward and faster than other high-resolution characterization approaches.
Crystalline Phase and Orientation Mapping
Crystalline phase and orientation mapping is extremely important to the development of engineering materials. Besides the method’s quotidian use for characterizing polycrystalline materials, it plays a decisive role in differentiating compounds that may present allotropes or polymorphs.
While the diffraction map implementation on the TEM enhances the spatial resolution in comparison to its SEM counterpart EBSD down to the few-nanometer scale, 4D STEM acquisition with a pixelated detector delivers improved yield and sensitivity when mapping crystalline domains and conducting relative orientation/texture analyses.
Orientation Mapping applied to an AlN polycrystalline sample with a columnar structure. (a) and (b) A virtual BF image and a sample localized diffraction pattern, respectively, referring to AlN along a  zone axis. (c) and (d) Orientation Mapping, split via in-plane rotation and out-of-plane tilt with respect to the main columns axis. Acquired with DECTRIS ARINA. Acknowledgment: Erdmann Spiecker and Mingjian Wu (FAU).