From Acquisition to Analysis—Streamlined 5D STEM Workflows for In Situ Experiments

4D STEM is a powerful technique for capturing detailed structural data. With advancements in camera and detector technology, rapid data acquisition is now possible. However, achieving optimal field of view and resolution in both real and reciprocal space requires acquiring large data volumes. Introducing the fifth dimension of time to capture a series of 4D data cubes during in-situ experiments further increases these volumes [1]. The mostly manual recording and processing of continuous 4D STEM (5D) datasets limited its adoption to a few specialized labs in the past.

In 2023, Gatan introduced software tools to streamline the continuous collection and processing of 5D in-situ 4D STEM datasets. Leveraging our eaSI™ technology, these tools will enhance the accessibility and utilization of 4D STEM for in-situ experiments. With DigitalMicrograph^® version 3.6.0 and a suitable license, a user can continuously capture 4D STEM data cubes, automatically saving data to disk after each scan until the user stops the recording. During acquisition, a virtual image is displayed in real time, allowing users to explore incoming data using our picker tool.

Post-acquisition, the 5D in-situ data can be immediately played back using the same free In-Situ Player tool Gatan provides for playing back in-situ TEM and STEM videos [2]. If a compatible in-situ holder from Gatan, DENSsolutions, or Protochips was connected to DigitalMicrograph, then holder metadata is automatically displayed and synchronized by the In-Situ Player also. Rapid alignment of diffraction pattern centers in reciprocal space and 4D cubes in real space can be performed over the entire 5D dataset using existing tools in DigitalMicrograph. Flexible virtual aperture maps can be dynamically computed from the current 4D cube or over the entire dataset.

Extracting relevant information from a 5D dataset requires a tailored approach dependent on sample characteristics, acquisition parameters, and desired outputs. Thus, flexibility in analysis and visualization is crucial. Incorporating Python scripts into DigitalMicrograph enables highly customized workflows for 5D datasets, leveraging existing open-source analysis packages.

This work demonstrates the workflow, ease of use, and speed of the new 5D STEM data collection, processing, and visualization tools in DigitalMicrograph. We also showcase Python processing of in-situ datasets, utilizing both custom techniques and existing routines from py4DSTEM [3]. The Python scripts operate on open datasets, enabling interactive use of built-in tools in DigitalMicrograph and producing results that can be played back by the In-Situ Player. These results are synchronized with raw data and in-situ holder data when any of them are played back, facilitating a more comprehensive understanding of analysis results in relation to in-situ conditions and raw data. Our workflows enable everything from acquisition to analysis in just a few hours on the same PC at the microscope.

[1] Gammer, C. et al. Local and transient nanoscale strain mapping during in situ deformation. Applied Physics Letters 109, 081906 (2016).
[2] In-situ Data Processing with Gatan Microscopy Suite 3.4 https://www.youtube.com/playlist?list=PL_kL-ZJRE__g98nJseGIoksAU68yBbljt (Accessed Feb 7 2023)
[3] Savitzky, B. H. et al. py4DSTEM: A Software Package for Four-Dimensional Scanning Transmission Electron Microscopy Data Analysis. Microsc Microanal 27, 712–743 (2021).