Total Scattering Analysis of Disordered Nanosheet Materials

Abstract

Two dimensional materials are of increasing interest as building blocks for functional coatings, catalysts, and electrochemical devices. While increasingly sophisticated processing routes have been designed to obtain high-quality exfoliated nanosheets and controlled, self-assembled mesostructures, structural characterization of these materials remains challenging. This work presents a novel method of analyzing pair distribution function (PDF) data for disordered nanosheet ensembles, where supercell stacking models are used to infer atom correlations over as much as 50 Å. Hierarchical models are used to reduce the parameter space of the refined model and help eliminate strongly correlated parameters.

Three data sets for restacked nanosheet assemblies with stacking disorder are analyzed using these methods: simulated data for graphene-like layers, experimental data for ~1 nm thick perovskite layers, and and experimental data for highly defective δ-MnO2 layers. In each case, the sensitivity of the PDF to the real-space distribution of layer positions is demonstrated by exploring the fit residual as a function of stacking vectors. The refined models demonstrate that nanosheets tend towards local interlayer ordering, which is hypothesized to be driven by the electrostatic potential of the layer surfaces. Correctly accounting for interlayer atom correlations permits more accurate refinement of local structural details including local structure perturbations and defect site occupancies. In the δ-MnO2 nanosheet material, the new modeling approach identified 14% Mn vacancies while application of 3D periodic crystalline models to the <7 Å PDF region suggests a 25% vacancy concentration. In contrast, the perovskite nanosheet material is demonstrated to exhibit almost negligible structural relaxation in contrast with the bulk crystalline material from which it is derived.