Work

Assessing Critical Dose for Beam-sensitive Samples using Low-dose Counted in-situ Video

Year: 2023
Type: article
Abstract: While transmission electron microscopy (TEM) can elucidate the structure of nanomaterials from the micro to the atomic scale, many functional materials are beam sensitive. This has made TEM characterization difficult or even impossible for some materials. The definition of “beam-sensitive” can vary greatly depending on who defines it and at what resolution the sample is observed. For the single-particle cryo-electron microscopy (cryo-EM) community, an accumulated dose of just tens of electrons per angstrom squared (e-/Å2) is enough to destroy a sample, but atomic resolution imaging of an individual protein is not expected [1,2]. For researchers studying materials at the atomic scale, a total dose of several thousand e-/Å2 may be considered low-dose, especially if the dose rate is low [3]. In contrast, for traditional atomic resolution imaging of robust TEM samples, the total dose may be on the order of 104-105 e-/Å2. Techniques that capture more than one image, like tomography, 4D STEM, or in-situ experiments, also tend to require larger electron doses. Whatever the magnitude of the critical dose above which a structure is likely to be significantly damaged, observation of the pristine material without damage is desired. Recently developed counting direct detection cameras can make observation of the pristine structure possible. Cameras like Gatan’s K2® and K3®, recent ThermoFisher Scientific Falcon cameras, and the Direct Electron Apollo can count each electron individually, significantly reducing readout noise and making it possible to operate the microscope at a very low dose rate on the sample. The exact range varies by camera and by the TEM magnification. In 2022 Gatan released two new counting cameras, the Alpine™, and the Metro™, optimized for low-dose, low-kV data collection. While the Alpine is specifically designed for cryo-EM applications, the Metro is intended for materials science research and exhibits excellent performance, whether capturing imaging, diffraction, or in-situ data. To prove that the pristine structure has been observed for a given material, one should observe the sample at a dose rate at which damage is expected to be slow and monitor the structural changes which are observed as a function of time. Ideally, these changes should be quantified. Since the Metro camera has built-in in-situ capability, continuous videos can be captured. This, combined with the drift correction capabilities of DigitalMicrograph®, makes observation and quantification of structural changes much easier. We can also observe the sample at a very low dose rate over a long period of time or even change the dose rate during a single video acquisition to quantify the damage accumulation. In this work, we will show a metal-organic framework (MOF) sample irradiated with a very low dose rate of ∼0.77 e-/Å2/s followed by 25 e-/Å2/s. The structural degradation is clearly observed at the higher dose rate in the in-situ video captured with the Metro camera. Several frames from this video are shown in Fig. 1 A-E. The damage was then quantified by measuring the changing intensity of lattice spots in the fast Fourier transforms of video frames using a radial profile, seen in Fig. 1 F-G. This revealed that damage at the very low dose rate was negligible. Summing all the drift-corrected frames from the low dose rate section of the video results in a high signal-to-noise image of the pristine structure, shown in Fig. 2 [4]. Frames from the in-situ video and corresponding FFTs. A-E) Left: Full frames and a smaller region (blue) from a drift-corrected in-situ video of a MOF sample under electron beam irradiation. Each frame here is a sum over 5 seconds of the original data, which was captured at five fps. A-E) Right: FFTs from these summed frames. F) A series of radial profiles computed from the FFTs of all frames in the video. The abrupt change in the contrast occurs at the time when the dose rate was increased. Arrows indicate the scale and correspond to the arrows in the FFTs of A and E. G) A plot produced by summing the radial profiles in F over all frequencies, yielding a graph of the lattice fringe visibility over time. The sum of drift-corrected frames from the first 65 seconds of the counted dataset shows the pristine structure of this MOF nanoparticle. The dose rate was 0.77 e-/Å2/s, and the total dose contributing to the displayed image is 50 e-/Å2.
Language: English
Open Access status: bronze
OpenAlex ID: W4385093946
FWCI:
Cites: 2
Cited by:
Related to: 10