Abstract text (incl. figure legends and references)
Introduction 3D Electron Diffraction (3DED), become the "low dose" technique for crystal structure determination of nanocrystals and a wide range of materials. The criteria for a successful crystal structure solution is data completeness while a sufficient refinement is data quality dependent.[1] Several methods for 3DED data acquisition using different hardware have been established, but in order to spread the usage of 3DED a standardization is needed to overcome two major challenges intrinsic in the hardware used.[2]
Objectives 3DED needs to become user friendly thanks to automatized data acquisition and diffractometer alignment. Secondly, the uncertainties intrinsically present in our microscopes need to be dumped which make it difficult to measure crystals with domains less than 100nm requiring smaller beam sizes. With these improvements, 3DED will be accessible to a wider community and will show its true potential compared to XRD analysis.
Methods The methods currently available use two major approaches, Nano-Beam Electron Diffraction (NBED) and Selected Area Electron Diffraction (SAED). SAED has several advantages, such as beam size down up to 200nm, small crystal shift during the tilt series. Combined with Instamatic, these features enable rapid dissemination of SAED, especially to JEOL users.[3] However, NBED can achieve beam sizes down to a few nm, reduce image dose rates up to 40-fold in STEM images, and use HAADF to generate Z-image contrast. It often shows higher quality datasets compared to SAED and allows multiple datasets acquisition from the same crystal, due to localized beam damage.[4] The concepts of the NBED technique are implemented in Fast-Automated Diffraction Tomography (Fast-ADT), a user-friendly DigitalMicrograph© (DM) script for 3DED acquisition and tracking.[5]
Results Here we report our improvement to the 3DED technique. Fast-ADT has been translated and implemented from DM© in Python. The script is based on the Temscript package, allowing use for the majority of the FEI microscopes and future expansion to JEOL microscopes using PyJEM.[6] Fast-ADT allows TEM/STEM SAED and NBED in Precession ED Tomography (PEDT) and Continuous Rotation ED (CRED). This way, FEI machines could use one universal tool for conventional 3DED experiment, with the potential to be deployed and customized to design its personal 3DED experiments. To enhance the 3DED automation, an automatic eucentric height tool is implemented in Fast-ADT, which performs the Z control of the target crystal in a more precise and reproducible way. This is essential for NBED, which plays a role to counterbalance the target drift during data collection and is the major responsible of a 3DED data collection failure.
Conclusion With these improvements, one of the last challenges that could be solved by software is the design of a more robust 3DED crystal tracking routine. This implementation is important for NBED because it is easier to lose the target during the tilt series than in SAED, where the beam size is higher. In conclusion, Fast-ADT has the potential to be used more efficiently and be spread easier to all the 3DED labs as a complete package for 3DED data acquisition.
[1]https://doi.org/10.1107/S2052520619006711
[2]https://doi.org/10.1021/acscentsci.9b00394
[3]https://doi.org/10.1107/S2052252519007681
[4]https://doi.org/10.1557/PROC-1184-GG01-05
[5]https://doi.org/10.1016/j.ultramic.2020.112951
[6]https://github.com/niermann/temscript