| 1. |
- Anker, Andy S., et al.
(author)
-
Extracting structural motifs from pair distribution function data of nanostructures using explainable machine learning
- 2022
-
In: npj Computational Materials. - : Springer Science and Business Media LLC. - 2057-3960. ; 8:1
-
Journal article (peer-reviewed)abstract
- Characterization of material structure with X-ray or neutron scattering using e.g. Pair Distribution Function (PDF) analysis most often rely on refining a structure model against an experimental dataset. However, identifying a suitable model is often a bottleneck. Recently, automated approaches have made it possible to test thousands of models for each dataset, but these methods are computationally expensive and analysing the output, i.e. extracting structural information from the resulting fits in a meaningful way, is challenging. Our Machine Learning based Motif Extractor (ML-MotEx) trains an ML algorithm on thousands of fits, and uses SHAP (SHapley Additive exPlanation) values to identify which model features are important for the fit quality. We use the method for 4 different chemical systems, including disordered nanomaterials and clusters. ML-MotEx opens for a type of modelling where each feature in a model is assigned an importance value for the fit quality based on explainable ML.
|
|
| 2. |
- Juelsholt, Mikkel, et al.
(author)
-
Size-induced amorphous structure in tungsten oxide nanoparticles
- 2021
-
In: Nanoscale. - : Royal Society of Chemistry (RSC). - 2040-3364 .- 2040-3372. ; 13:47, s. 20144-20156
-
Journal article (peer-reviewed)abstract
- The properties of functional materials are intrinsically linked to their atomic structure. When going to the nanoscale, size-induced structural changes in atomic structure often occur, however these are rarely well-understood. Here, we systematically investigate the atomic structure of tungsten oxide nanoparticles as a function of the nanoparticle size and observe drastic changes when the particles are smaller than 5 nm, where the particles are amorphous. The tungsten oxide nanoparticles are synthesized by thermal decomposition of ammonium metatungstate hydrate in oleylamine and by varying the ammonium metatungstate hydrate concentration, the nanoparticle size, shape and structure can be controlled. At low concentrations, nanoparticles with a diameter of 2-4 nm form and adopt an amorphous structure that locally resembles the structure of polyoxometalate clusters. When the concentration is increased the nanoparticles become elongated and form nanocrystalline rods up to 50 nm in length. The study thus reveals a size-dependent amorphous structure when going to the nanoscale and provides further knowledge on how metal oxide crystal structures change at extreme length scales.
|
|
