Search for: All records

This article reports the study of algorithms for non-negative matrix factorization (NMF) in various applications involving smoothly varying data such as time or temperature series diffraction data on a dense grid of points. Utilizing the continual nature of the data, a fast two-stage algorithm is developed for highly efficient and accurate NMF. In the first stage, an alternating non-negative least-squares framework is used in combination with the active set method with a warm-start strategy for the solution of subproblems. In the second stage, an interior point method is adopted to accelerate the local convergence. The convergence of the proposed algorithm is proved. The new algorithm is compared with some existing algorithms in benchmark tests using both real-world data and synthetic data. The results demonstrate the advantage of the algorithm in finding high-precision solutions.

 

We present a deep learning algorithm, DeepStruc, that can solve a simple nanoparticle structure directly from an experimental Pair Distribution Function (PDF) by using a conditional variational autoencoder.

 

Structural modelling of octahedral tilts in perovskites is typically carried out using the symmetry constraints of the resulting space group. In most cases, this introduces more degrees of freedom than those strictly necessary to describe only the octahedral tilts. It can therefore be a challenge to disentangle the octahedral tilts from other structural distortions such as cation displacements and octahedral distortions. This paper reports the development of constraints for modelling pure octahedral tilts and implementation of the constraints in diffpy-CMI , a powerful package to analyse pair distribution function (PDF) data. The model in the program allows features in the PDF that come from rigid tilts to be separated from non-rigid relaxations, providing an intuitive picture of the tilting. The model has many fewer refinable variables than the unconstrained space group fits and provides robust and stable refinements of the tilt components. It further demonstrates the use of the model on the canonical tilted perovskite CaTiO 3 which has the known Glazer tilt system α + β − β − . The Glazer model fits comparably to the corresponding space-group model Pnma below r = 14 Å and becomes progressively worse than the space-group model at higher r due to non-rigid distortions in the real material. 

Machine learning models based on convolutional neural networks have been used for predicting space groups of crystal structures from their atomic pair distribution function (PDF). However, the PDFs used to train the model are calculated using a fixed set of parameters that reflect specific experimental conditions, and the accuracy of the model when given PDFs generated with different choices of these parameters is unknown. In this work, the results of the top-1 accuracy and top-6 accuracy are robust when applied to PDFs of different choices of experimental parameters r max , Q max , Q damp and atomic displacement parameters. 

Abstract Deep learning (DL) is one of the fastest-growing topics in materials data science, with rapidly emerging applications spanning atomistic, image-based, spectral, and textual data modalities. DL allows analysis of unstructured data and automated identification of features. The recent development of large materials databases has fueled the application of DL methods in atomistic prediction in particular. In contrast, advances in image and spectral data have largely leveraged synthetic data enabled by high-quality forward models as well as by generative unsupervised DL methods. In this article, we present a high-level overview of deep learning methods followed by a detailed discussion of recent developments of deep learning in atomistic simulation, materials imaging, spectral analysis, and natural language processing. For each modality we discuss applications involving both theoretical and experimental data, typical modeling approaches with their strengths and limitations, and relevant publicly available software and datasets. We conclude the review with a discussion of recent cross-cutting work related to uncertainty quantification in this field and a brief perspective on limitations, challenges, and potential growth areas for DL methods in materials science. 

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. OurMachineLearning basedMotifExtractor (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.

 

The development of new nanomaterials for energy technologies is dependent on understanding the intricate relation between material properties and atomic structure. It is, therefore, crucial to be able to routinely characterise the atomic structure in nanomaterials, and a promising method for this task is Pair Distribution Function (PDF) analysis. The PDF can be obtained through Fourier transformation of x-ray total scattering data, and represents a histogram of all interatomic distances in the sample. Going from the distance information in the PDF to a chemical structure is an unassigned distance geometry problem (uDGP), and solving this is often the bottleneck in nanostructure analysis. In this work, we propose to use a Conditional Variational Autoencoder (CVAE) to automatically solve the uDGP to obtain valid chemical structures from PDFs. We use a simple model system of hypothetical mono-metallic nanoparticles containing up to 100 atoms in the face centered cubic (FCC) structure as a proof of concept. The model is trained to predict the assigned distance matrix (aDM) from a simulated PDF of the structure as the conditional input. We introduce a novel representation of structures by projecting them inside a unit sphere and adding additional anchor points or satellites to help in the reconstruction of the chemical structure. The performance of the CVAE model is compared to a Deterministic Autoencoder (DAE) showing that both models are able to solve the uDGP reasonably well. We further show that the CVAE learns a structured and meaningful latent embedding space which can be used to predict new chemical structures. 

SASPDF, a method for characterizing the structure of nanoparticle assemblies (NPAs), is presented. The method is an extension of the atomic pair distribution function (PDF) analysis to the small-angle scattering (SAS) regime. The PDFgetS3 software package for computing the PDF from SAS data is also presented. An application of the SASPDF method to characterize structures of representative NPA samples with different levels of structural order is then demonstrated. The SASPDF method quantitatively yields information such as structure, disorder and crystallite sizes of ordered NPA samples. The method was also used to successfully model the data from a disordered NPA sample. The SASPDF method offers the possibility of more quantitative characterizations of NPA structures for a wide class of samples.