Phase quantification using deep neural network processing of XRD patterns

Mineral identification and quantification are key to the understanding and, hence, the capacity to predict material properties. The method of choice for mineral quantification is powder X-ray diffraction (XRD), generally using a Rietveld refinement approach. However, a successful Rietveld refinement requires preliminary identification of the phases that make up the sample. This is generally carried out manually, and this task becomes extremely long or virtually impossible in the case of very large datasets such as those from synchrotron X-ray diffraction computed tomography. To circumvent this issue, this article proposes a novel neural network (NN) method for automating phase identification and quantification. An XRD pattern calculation code was used to generate large datasets of synthetic data that are used to train the NN. This approach offers significant advantages, including the ability to construct databases with a substantial number of XRD patterns and the introduction of extensive variability into these patterns. To enhance the performance of the NN, a specifically designed loss function for proportion inference was employed during the training process, offering improved efficiency and stability compared with traditional functions. The NN, trained exclusively with synthetic data, proved its ability to identify and quantify mineral phases on synthetic and real XRD patterns. Trained NN errors were equal to 0.5% for phase quantification on the synthetic test set, and 6% on the experimental data, in a system containing four phases of contrasting crystal structures (calcite, gibbsite, dolomite and hematite). The proposed method is freely available on GitHub and allows for major advances since it can be applied to any dataset, regardless of the mineral phases present.

Crystallographic phase identifier of a convolutional self-attention neural network (CPICANN) on powder diffraction patterns

Spectroscopic data, particularly diffraction data, are essential for materials characterization due to their comprehensive crystallographic information. The current crystallographic phase identification, however, is very time consuming. To address this challenge, we have developed a real-time crystallographic phase identifier based on a convolutional self-attention neural network (CPICANN). Trained on 692 190 simulated powder X-ray diffraction (XRD) patterns from 23 073 distinct inorganic crystallographic information files, CPICANN demonstrates superior phase-identification power. Single-phase identification on simulated XRD patterns yields 98.5 and 87.5% accuracies with and without elemental information, respectively, outperforming JADE software (68.2 and 38.7%, respectively). Bi-phase identification on simulated XRD patterns achieves 84.2 and 51.5% accuracies, respectively. In experimental settings, CPICANN achieves an 80% identification accuracy, surpassing JADE software (61%). Integration of CPICANN into XRD refinement software will significantly advance the cutting-edge technology in XRD materials characterization.

Neural networks for rapid phase quantification of cultural heritage X-ray powder diffraction data

Recent developments in synchrotron radiation facilities have increased the amount of data generated during acquisitions considerably, requiring fast and efficient data processing techniques. Here, the application of dense neural networks (DNNs) to data treatment of X-ray diffraction computed tomography (XRD-CT) experiments is presented. Processing involves mapping the phases in a tomographic slice by predicting the phase fraction in each individual pixel. DNNs were trained on sets of calculated XRD patterns generated using a Python algorithm developed in-house. An initial Rietveld refinement of the tomographic slice sum pattern provides additional information (peak widths and integrated intensities for each phase) to improve the generation of simulated patterns and make them closer to real data. A grid search was used to optimize the network architecture and demonstrated that a single fully connected dense layer was sufficient to accurately determine phase proportions. This DNN was used on the XRD-CT acquisition of a mock-up and a historical sample of highly heterogeneous multi-layered decoration of a late medieval statue, called 'applied brocade'. The phase maps predicted by the DNN were in good agreement with other methods, such as non-negative matrix factorization and serial Rietveld refinements performed with TOPAS, and outperformed them in terms of speed and efficiency. The method was evaluated by regenerating experimental patterns from predictions and using the R-weighted profile as the agreement factor. This assessment allowed us to confirm the accuracy of the results.

Crystal Structure Assignment for Unknown Compounds from X-ray Diffraction Patterns with Deep Learning

Determining the structures of previously unseen compounds from experimental characterizations is a crucial part of materials science. It requires a step of searching for the structure type that conforms to the lattice of the unknown compound, which enables the pattern matching process for characterization data, such as X-ray diffraction (XRD) patterns. However, this procedure typically places a high demand on domain expertise, thus creating an obstacle for computer-driven automation. Here, we address this challenge by leveraging a deep-learning model composed of a union of convolutional residual neural networks. The accuracy of the model is demonstrated on a dataset of over 60,000 different compounds for 100 structure types, and additional categories can be integrated without the need to retrain the existing networks. We also unravel the operation of the deep-learning black box and highlight the way in which the resemblance between the unknown compound and a structure type is quantified based on both local and global characteristics in XRD patterns. This computational tool opens new avenues for automating structure analysis on materials unearthed in high-throughput experimentation.

Fast extraction of three-dimensional nanofiber orientation from WAXD patterns using machine learning

Structural disclosure of biological materials can help our understanding of design disciplines in nature and inspire research for artificial materials. Synchrotron microfocus X-ray diffraction is one of the main techniques for characterizing hierarchically structured biological materials, especially the 3D orientation distribution of their interpenetrating nanofiber networks. However, extraction of 3D fiber orientation from X-ray patterns is still carried out by iterative parametric fitting, with disadvantages of time consumption and demand for expertise and initial parameter estimates. When faced with high-throughput experiments, existing analysis methods cannot meet the real time analysis challenges. In this work, using the assumption that the X-ray illuminated volume is dominated by two groups of nanofibers in a gradient biological composite, a machine-learning based method is proposed for fast and automatic fiber orientation metrics prediction from synchrotron X-ray micro-focused diffraction data. The simulated data were corrupted in the training procedure to guarantee the prediction ability of the trained machine-learning algorithm in real-world experimental data predictions. Label transformation was used to resolve the jump discontinuity problem when predicting angle parameters. The proposed method shows promise for application in the automatic data-processing pipeline for fast analysis of the vast data generated from multiscale diffraction-based tomography characterization of textured biomaterials.

Interpretable Deep Learning Model for Analyzing the Relationship between the Electronic Structure and Chemisorption Property

The use of machine learning (ML) is exploding in materials science as a result of its high predictive performance of material properties. Tremendous trainable parameters are required to build an outperforming predictive model, which makes it impossible to retrace how the model predicts well. However, it is necessary to develop a ML model that can extract human-understandable knowledge while maintaining performance for a universal application to materials science. In this study, we developed a deep learning model that can interpret the correlation between surface electronic density of states (DOSs) of materials and their chemisorption property using the attention mechanism that provides which part of DOS is important to predict adsorption energies. The developed model constructs the well-known d-band center theory without any prior knowledge. This work shows that human-interpretable knowledge can be extracted from complex ML models.