Modern environmental Transmission Electron Microscopes (ETEM) enables chemical reactions to be directly observed with new perspectives in the operando characterization of nano-materials. However, morphological features are essentially missing in 2-dimensional observations, thus nano-tomography under environmental conditions is a new promising challenge. Obviously, the essential condition to achieve this goal is to run fast tilt series acquisitions as compared to the kinetics of the reactions which are followed in situ in the microscope. This contribution will show that such experiments are possible by comparing the volumes respectively obtained from a classic or a fast tilt series acquisition in the bright field mode.

Firstly, simulations were performed on ghost models in order to appreciate the influence of the goniometer rotation speed during image acquisition on quality of images (sharpness and blurring effects). A typical micrograph of a nano-object, e.g. metallic nanoparticles encaged into mesoporous silicalites, was used to reconstruct a 2D model. The 1D projections were calculated according to different conditions intending to reproduce the effects of a continuous tilt during the acquisition. Figure 1 a-c) show the models at zero tilt projected perpendicularly to the tilt axis marked by a cross (the vertical direction is the projection direction); in a), a fixed image is shown as obtained at a given rotation; it is compared to images simulated by integrating a blur effect to a rotation of 3° during the acquisition, either whit a centered (b) or not centered rotation axis (c). To give an order of magnitude, a 120° rotation performed in 1 minute with acquisition of images every second without interrupting the rotation leads to an angular blur of only 2° in each image. From the 1D projection series (not shown here), 2D reconstructions were calculated using the simple Weighted-Back Projection (WBP) algorithm. Results from fig. 1 d-e) show that, at least in the case of nanoparticles with strong absorption contrast as presented here, the tomograms obtained from the blurred series are not significantly different from the constructed volume obtained from the conventional step-by-step acquisition scheme.

In a second step, we performed experimental nano-tomography experiments on Pd/Al₂O₃ samples deposited on holey carbon grids. Volume reconstructions shown in Fig.2 were obtained from the same object using two bright field tilt series acquired in a FEI Titan-ETEM microscope operated at 300 kV and equipped with a dedicated Fischione high-tilt sample holder. The first one was acquired through a classical step by step tilt series acquisition from 74° to + 66° with a step of 2° in mode Saxton within 45 minutes. The second one was recorded by ‘fast tomography’ in 150 seconds. From these data, a quantitative analysis of the Pd nanoparticles (NP) distribution and size was performed and reported in Fig. 3. Although differences obviously exist (especially, the fast tomography approach misses some of the NPs smaller than nominally 2 nm and tends to overestimate the size of the largest ones), it can be concluded that acquisitions of tilting series in very short times of the order of one minute, or even less, represent a promising way to provide 3D information on samples studied under dynamic gas and temperature conditions such as typically nano-catalysts studied in an Environmental TEM. This fast tomography approach can also be of a great interest for beam sensitive samples where the material is generally not able to bear a long exposure to the electron beam without any specific and sometimes hazardous pre-treatment or preparation.

Acknowledgements

Thanks are due to CLYM (Consortium Lyon – St-Etienne de Microscopie, www.clym.fr) for the access to the microscope funded by the Region Rhône-Alpes, the CNRS and the ‘GrandLyon’.

Acknowledgments are also due to BQR project SEE3D granted by Insa-Lyon, ANR project 3DClean, Labex iMUST and IFP Energies Nouvelles for the financial support.

Figures:

Fig. 1: Tilting tomography reconstructions of 2D images. a): (x,z) view of the 'volume' at zero tilt without blur; b-c) same as a) with a integrated rotation blur of 3°; the cross symbol shows the rotation axis. d-f): corresponding WBP reconstructions from 1D projections along z from -60 to 60°, step = 3°.

Fig. 2: Reconstructed volumes of an alumina grain supporting Pd nanoparticles (shown in red). Right: conventional bright field tomography, left: fast tomography.

Fig. 3: Pd particles size distribution obtained by classic tomography and fast tomography (average size respectively equal to 4.1 nm and 3.9 nm).

« Back to The 16th European Microscopy Congress 2016

EMC Abstracts - https://emc-proceedings.com/abstract/environmental-transmission-electron-tomography-fast-3d-analysis-of-nano-materials/