Improving Electron Pair Distribution Function (ePDF) for Nanomaterials—The Case of Iron Oxide Nanoparticles

Nanomaterials’ properties are strongly related to their crystallographic atomic arrangement. Methods to characterize their atomic structure are fundamental to understanding materials' properties and, consequently, controlling their properties and achieving their full performance. However, classical X-ray diffraction (XRD) methods fail to elucidate the nanostructures due to the elevated peak broadening, a challenge that is known as the nanostructure problem. Pair Distribution Function (PDF) analysis obtained by neutrons and X-ray scattering (synchrotron) have been used to study nanomaterials with outstanding performance, but the access to synchrotron data is limited to the general nanoscience researchers. Electrons have become an alternative source as, nowadays, a transmission electron microscope (TEM) is certainly more accessible compared to synchrotrons. Electrons also have higher scattering power offering the advantage of using a small amount of sample , time-efficient data acquisition, and lower costs.¹ On the opposite side, electrons have a higher probability of multiple scattering and inelastic scattering leading to a harder data processing to collect the total scattering in the kinematical regime. Then, advances in implementing the PDF analysis using electron diffraction (ePDF) would be of great importance.¹ ePDF is a technique capable of analyzing materials structure directly in the real space with the <i>G(r)</i> profile leading to the probability of occurrence of any pair of atoms within the material by their interatomic distances <i>r</i>.¹ Also, ePDF can reveal all nanostructure features, including bulk crystalline region to the surface, which frequently possesses higher atomic disorder.² In this work, methods to improve ePDF <i>G(r)</i> profile were implemented by comparing the experimental scattering profile <i>I(Q)</i>_exp, structure factor <i>S(Q)</i>_exp, and the PDF <i>G(r)</i>_exp with calculated <i>I</i>(<i>Q</i>)_calc, <i>S</i>(<i>Q</i>)_calc, and <i>G(r)</i>_calc for spherical nanoparticles. Size-controlled iron oxide nanoparticles were synthesized to be used as a case study for ePDF data acquisition and processing toward quantitative <i>G(r) </i>data analysis. The electron diffraction data acquisition was optimized by collecting the scattering of a self-sustained nanoparticle superlattice membrane to avoid a carbon background. The scattering from the organic ligands (oleic acid) was also removed from the <i>I(Q)</i>_exp to acquire only the electron scattering from the nanomaterial. By applying a frequency filter, the contribution of possible multiple and inelastic scattering background on <i>S(Q)</i>_exp was removed. Finally, the <i>G(r)</i>_exp was multiplied by a constant to achieve approximate the signal with the <i>G(r)</i>_calc. The final ePDF <i>G(r)</i> was compared to a nanostructure model using a structure refinement method in PDFgui software. Several iron oxide nanoparticles were used to test the proposed method, from the as-synthesized magnetite to oxidized maghemite nanoparticles. Excellent structure refinement was achieved with R_w below 14% showing that our new method of ePDF data acquisition and processing can be used as a quantitative atomic structure method for the short-range atomic ordering and the medium-range ordering (&lt; 10 nm) present in nanoparticles. combining regular TEM analysis like morphology (size, size distribution, and shape), high-resolution, and chemical mapping, the ePDF atomic structure analysis adds to TEM a perfect crystallographic characterization tool to solve the nanostructure problem.<br/>The authors would like to acknowledge FAPESP (Grants 2018/05159-1, 2021/03321-9) and CNPq (409787/2021-3).<br/>[1] Souza Junior, J.B. et al. <b>Matter</b>. 2021, 4 (2), 441-460.<br/>[2] Souza Junior, J.B. et. al. <b>J. Phys. Chem. Lett.</b> 2020, 11 (4), 1564-1569.