Powder Nano-Beam Diffraction in Scanning Electron Microscopy: Possibilities and Limitations for Applications

We have recently introduced a novel scanning electron microscopy (SEM) method which enables fast and simple analysis of nanocrystalline materials using a 2D pixelated detector of transmitted electrons. This is called 4D-STEM, as we obtain 2D information for each pixel of the 2D scanning array. On the one hand, the 2D-STEM detectors can be used routinely to collect a high number of electron diffraction patterns from individual nanocrystals and locations, but on the other hand, the individual 4D-STEM diffractograms are challenging to analyze due to the random orientation of nanocrystalline material. In our approach, we present that all individual diffractograms which show randomly oriented diffraction spots from a few nanocrystals are combined into one composite diffraction pattern. Consequently, this pattern shows diffraction rings which are typical for polycrystalline/powder materials [1]. The final powder diffraction pattern can be analyzed by employing standard programs for TEM/SAED (Selected-Area Electron Diffraction), such as Process Diffraction [2]. We called our new method 4D-STEM/PNBD (Powder NanoBeam Diffraction) [3]. The method was designed with a focus on creating a robust and easy-to-use crystal identification method for both sparsely dispersed and embedded particles, respectively. The 2D-STEM detectors for STEM-in-SEM have been commercialized recently [4]. The acquired 4D data of electron diffraction patterns from individual nanocrystals can be combined into one composite powder diffraction pattern (Fig. 1). The original version of the 4D-STEM/PNBD method [1], which suffered from low resolution, was improved in three essential areas: (i) an optimized data collection protocol enables the experimental determination of the point spread function (PSF) of the primary electron beam, (ii) an improved data processing combines entropy-based filtering of the whole dataset with a PSF-deconvolution of the individual 2D diffractograms, and (iii) completely re-written software automates all calculations and requires just a minimal user input. The user-friendly freeware packages were written in Python [5] and MATLAB [6] environments. Both programs were designed for simple, routine, and efficient everyday use with minimal user input. In the present study, we report on the practical limitations of this method in relation to different background thicknesses of the supporting carbon film. Moreover, we propose selected biological applications of this method, where uniqueness in the identification of nanoparticles is presented [7]. Principle of 4D-STEM/PNBD method. The left part of the image shows how the scanning electron beam penetrates through the sample and forms a set of nanobeam diffraction (NBD) patterns. The right part of the image shows how the individual NBD patterns can be combined into one powder nanobeam diffraction pattern (PNBD; with diffraction rings typical of the diffraction on polycrystalline samples): (a) PNBD obtained by the straightforward summation of all NBD patterns, (b) PNBD obtained by the summation of selected NBD patterns with high Shannon entropy values (high entropy indicated strongly diffracting crystals), (c) point spread function (PSF) calculated from low-entropy NBD patterns (the PSF represents the XY-spread of the primary beam), and (d) PNBD obtained by the summation of high-entropy NBD patterns individually preprocessed by PSF deconvolution. Reprinted from [3].