Advanced characterization of NP@MOF core-shell complexes using real-time electron tomography

Abstract text (incl. figure legends and references)

Metal-organic Frameworks (MOFs) are a diverse class of materials consisting of metal clusters and organic linkers arranged into a porous crystalline lattice. Due to their porosity, MOFs possess some of the largest surface areas of any known material. This is exploited for gas adsorption, catalysis and drug delivery. For drug delivery, a drug is incorporated within the MOF and release is initiated by stimuli induced MOF decomposition. Recent efforts have focused on incorporating stimuli responsive nanoparticles (NP) into the MOF, to induce thermal decomposition as a delivery mechanism. The morphology of these NP@MOF complexes is critical to their overall efficacy for drug delivery. [1]

Electron tomography (ET) is invaluable for characterizing their 3D morphology. However, ET requires long exposure times and high electron dosages which cause artefacts during imaging of NP@MOFs. Limiting beam exposure can be achieved by limiting the number of images acquired. Unfortunately, reducing the image number also induces undersampling artefacts. Hence, an optimum image number must be obtained. Optimization could be performed by comparing multiple 3D reconstructions acquired with variable image numbers, a time-consuming process. Furthermore, since each acquisition must be performed on a new sample due to beam exposure from the previous acquisition, reconstructions are not directly comparable and a microscopist must perform a qualitative judgement with uncertain reliability. [2]

Herein a protocol was propsed to estimate the optimum number of images from a single acquisition by combining a non-termination acquisition protocol, golden ratio scanning (grs) with real-time reconstruction. The protocol was tested using simulations and an experimentally acquired Au/Pd NP. Finally, the value of this procedure was highlighted by application to NP@MOF samples of ZIF-8 and NU-1000 containing Au nanoparticles. [3]

Fig. 1. In ET , images are collected (a) incrementally from the lowest angle to the highest angle. E.g. image 1 is collected at -70°, image 2 collected at -65° etc. Early termination results in a missing wedge. By collecting images using (b) grs, the ,issing wedge is filled mostly in the first images and subsequent images improve sampling resolution. Real time reconstruction is used to monitor reconstruction quality and terminate acquisition. For Au/Pd@ZIF-8 viewed along the xy, xz and yz planes, undersampled (10 images) and oversampled (27) reconstructions have artefacts (red) not apparent in the determined optimum (13 images).

References

[1] Wang, L., Zheng, M., and Xie, Z. (2018) Nanoscale metal–organic frameworks for drug delivery: a conventional platform with new promise. Journal of Materials Chemistry B, 6 (5), 707–717

[2] Vanrompay, H., Béché, A., Verbeeck, J., and Bals, S. (2019) Experimental Evaluation of Undersampling Schemes for Electron Tomography of Nanoparticles. Particle and Particle Systems Characterization, 36 (7), 1–8.

[3] This project receives funding from the European Union"s Horizon 2020 research and innovation programme grant agreement No 860942 (HEATNMOF).