NLSTEM: Non-local denoising for enhanced 4D-STEM pattern indexing

Abstract

4D-STEM-based orientation and phase mapping has enabled rapid microstructure quantification that can be directly combined with standard TEM- and STEM-based imaging modes. Typically, orientation mapping is coupled with beam precession (i.e. precession electron diffraction) to achieve high indexing rates, adding to the cost and often decreasing the spatial resolution of the approach. This paper introduces a new post processing approach modeled after the non-local pattern averaging and reindexing algorithm developed for the electron backscatter diffraction community, wherein post-collection, patterns are averaged using a distance similarity parameter. Results from Ni and Au thin films show that indexing rates can be significantly improved using this post-processing technique due to improved signal-to-noise ratios in the diffraction patterns. Interestingly, the highest indexing rates are achieved in samples heavily damaged via ion irradiation, suggesting that averaging over curved lattices further improves indexing rates.

Figures (4)

• Figure 1: 4D-STEM orientation maps of a nanocrystalline Ni thin film obtained from as-collected data and after NLSTEM processing. (a) As-collected IPF orientation map. Inset in (a) is a representative raw convergent-beam electron diffraction pattern, illustrating the faint, noisy Bragg discs caused by the nanoscale grain mixture. (b) Corresponding IPF orientation map after NLSTEM processing of the same dataset, yielding a nearly fully indexed map. The enhanced diffraction pattern quality is shown in inset in (b)
• Figure 2: Effect of NLSTEM processing on 4D-STEM crystal orientation maps of irradiated Au thin films at 0, 1, and 5 dpa. Top row: as-collected indexing results; bottom row: after NLSTEM processing and re-indexing. The precent of reliably indexed data points is included with each map.
• Figure 3: Effect of NLSTEM processing on a representative diffraction pattern from 5 dpa irradiated Au. (a) Diffraction pattern prior to NLSTEM processing. (b) The same diffraction pattern after NLSTEM processing. (c) Intensity profile taken along the blue lines in (a) and (b) showing the increased signal to noise ratio after NLSTEM processing. (d) Segment of the intensity profile shown in boxed region in (c) highlighting minor diffraction peaks that can be identified after NLSTEM processing. A blue arrow is used to indicated one example of a faint peak only identifiable in the post-processed pattern.
• Figure 4: IPF maps of the same area in an annealed Ni thin film indexed using NPAR and NLSTEM under two pattern-collection step sizes: 10 nm (fine) and 20 nm (coarse). Dashed circles mark thin twin lamellae with thicknesses of approximately $\sim\!10$ nm (1) and $\sim\!20$ nm (2).