Electron pair distribution function (ePDF) analysis of micro- and nanoplastic particles

Abstract text (incl. figure legends and references)

Micro- and nanoplastics (MNPs) are considered a possible threat to microorganisms in the aquatic environment. Here, we show that total scattering intensity analysis of electron diffraction (ED) data measured by transmission electron microscopy (TEM), which yields the electron pair distribution function (ePDF),¹ is a feasible method for the characterization and identification of MNPs down to 100 nm.

Polyethylene (PE), polypropylene (PP), polyethylene terephthalate (PET), and polyamide microparticles, and polystyrene (PS) nanoparticles were examined by TEM and serial ED under low dose, non-cryogenic conditions. Reduced pair density functions (RDF) were calculated from the processed and background subtracted ED data and compared to simulated RDF. Nanobeam diffraction was used for the measurements of the PS and SiO₂ nanoparticles.

High quality serial ED suitable for PDF analysis were measured on polymer microparticles (Figure 1) and nanoparticles. The ePDF method yields a qualitative agreement of experimentally determined and simulated RDFs (Figure 2). For the nanoparticulate PS, a distinction from SiO₂ particles can be made (Figure 3), where the SiO₂ particles are representative of the natural inorganic matrix. By analyzing a time series of ED, we are able to show the dose dependent structural changes in the materials in the series of RDFs. Those changes are qualitatively different for the two different dose rates that were applied, which is evident in particular in the peak intensities that can be attributed to C-H-bonds and C-C-bonds.

We show that detailed structural information can be gained on highly beam sensitive polymer materials by following a low dose, ePDF approach without applying cryogenic or staining techniques. The systematic appearance of C-H pair distance peaks in our measurements exemplifies an advantage of the ePDF method over X-ray based PDF, which usually cannot gain information about hydrogen.

Figure 1 ED (top row) and TEM (bottom row) of cryo ball milled PE, PP, PET, and PA microparticles.

Figure 2 Reduced pair density functions derived from serial ED measured on cryo ball milled PE, PP, PET, and PA microplastic particles, respectively. A) Results of measurements taken with a dose rate of ≈ 45 e/(nm2s). B) 3D model structures employed in the RDF simulation. C) Results of measurements taken with a dose rate of a dose rate of ≈ 160 e/(nm² s). Each graph in A and C shows the experimental time series of RDFs, G(r) (color lines), the calculated RDF of a model structure, G(r) model (grey lines), and the elemental pair distribution function of the model (stacked bar charts). The color scale bar indicates the total exposure time at the end of acquisition of an ED frame.

Figure 3 Comparison of ePDF of 100 nm sized PS with 100 nm sized a-SiO₂ spheres. A) TEM of PS particles after ED measurements. B) First frame of serial ED measured on a PS sphere, C) Shadow image of the beam (diameter ~75 nm, indicated by the red dashed line) on the PS sphere (indicated by the white dashed line). D) TEM of a-SiO2 particles after ED measurements. E) First frame of serial ED measured on an a-SiO₂ sphere F) Shadow image of the beam (diameter ~75 nm) on the a-SiO₂ sphere. G) Plots of corresponding RDFs of PS (top) and SiO2 (bottom). H) Structural model of PS I) Structural model of SiO₂.

(1) Gorelik TE et al. Acta Cryst. 2019;75:532–49.