Early Detection of Nucleation Events From Solution in LC-TEM by Machine Learning

Abstract

To support the detection, recording, and analysis of nucleation events during in situ observations, we developed an early detection system for nucleation events observed using a liquid-cell transmission electron microscope. Detectability was achieved using the machine learning equivalent of detection by humans watching a video numerous times. The detection system was applied to the nucleation of sodium chloride crystals from a saturated acetone solution of sodium chlorate. Nanoparticles with a radius of more greater than 150 nm were detected in a viewing area of 12 μm × 12 μm by the detection system. The analysis of the change in the size of the growing particles as a function of time suggested that the crystal phase of the particles with a radius smaller than 400 nm differed from that of the crystals larger than 400 nm. Moreover, the use of machine learning enabled the detection of numerous nanometer sized nuclei. The nucleation rate estimated from the machine-learning-based detection was of the same order as that estimated from the detection using manual procedures.

Electron Microscopy Studies of Soft Nanomaterials

“…The contours were then used to calculate the etching kinetics, diffusion coefficient, interaction potential, and self-assembly kinetics of NPs. Another example of applying the object detection models is to achieve real-time detection of early nucleation events during liquid-phase TEM experiments . The study focused on observing the NaCl crystallization in liquid cell starting from NaClO 3 dissolved in acetone.…”

“…Another example of applying the object detection models is to achieve real-time detection of early nucleation events during liquid-phase TEM experiments. 635 The study focused on observing the NaCl crystallization in liquid cell starting from NaClO 3 dissolved in acetone. Upon beam illumination, the Cl − produced by ClO 3 − radiolysis crystallizes with Na + .…”

Molecular Mechanism of Organic Crystal Nucleation: A Perspective of Solution Chemistry and Polymorphism

“…Other attempts combined high-throughput experimental methods and the statistical analysis of measured induction times to explore the correlation of crystallization conditions [125,126], solvation and dimerization of solute molecules in nucleation assembly [127], molecular attachment frequency and interfacial energy [36,122], and intermolecular interaction energies [128] with nucleation kinetics. In addition, machine learning has recently been applied to the study of nucleation kinetics in some inorganic and colloidal nanoparticle systems to estimate the nucleation rate or the formation rate of precursors [129][130][131].…”

Possible embryos and precursors of crystalline nuclei of calcium carbonate observed by liquid-cell transmission electron microscopy

“…We have previously developed a low-dose-electron observation method that uses machine learning to reduce the effect of radiolysis by the electron beam, 25 together with a method for the early detection of nucleation events by machine learning that permits efficient observations on small solution volumes (or small observation areas). 26 On the basis of these previous studies, we have now developed a machine-learning program that uses a standard YOLOv5 algorithm (https://github.com/ultralytics/yolov5) to detect particles in a dataset of in situ TEM images. Annotation was performed using the open-source program, Label-Img (https://github.com/tzutalin/labelImg) for training and validation, we used different 40 s portions of a single 10 fps video recording.…”

“…The YOLO algorithm for object detection works effectively with TEM images. 26 We prepared training datasets for the detection of pre-nucleation particles or embryos. The images were extracted from HRTEM observations captured as video recordings.…”