Search NASA⌕ Search

Engineering topics

Demortiere, Arnaud

Publications and source records attributed to Demortiere, Arnaud.

Artificial neural network approach for multiphase segmentation of battery electrode nano-CT images

The segmentation of tomographic images of the battery electrode is a crucial processing step, which will have an additional impact on the results of material characterization and electrochemical simulation. However, manually labeling X-ray CT images (XCT) is time-consuming, and these XCT images are generally difficult to segment with histographical methods. We propose a deep learning approach with an asymmetrical depth encode-decoder convolutional neural network (CNN) for real-world battery material datasets. This network achieves high accuracy while requiring small amounts of labeled data and predicts a volume of billions voxel within few minutes. While applying supervised machine learning for segmenting real-world data, the ground truth is often absent. The results of segmentation are usually qualitatively justified by visual judgement. We try to unravel this fuzzy definition of segmentation quality by identifying the uncertainty due to the human bias diluted in the training data. Further CNN trainings using synthetic data show quantitative impact of such uncertainty on the determination of material’s properties. Nano-XCT datasets of various battery materials have been successfully segmented by training this neural network from scratch. We will also show that applying the transfer learning, which consists of reusing a well-trained network, can improve the accuracy of a similar dataset.

25 ENERGY STORAGE↗

Adorym: a multi-platform generic X-ray image reconstruction framework based on automatic differentiation

We describe and demonstrate an optimization-based X-ray image reconstruction framework called Adorym. Our framework provides a generic forward model, allowing one code framework to be used for a wide range of imaging methods ranging from near-field holography to fly-scan ptychographic tomography. By using automatic differentiation for optimization, Adorym has the flexibility to refine experimental parameters including probe positions, multiple hologram alignment, and object tilts. It is written with strong support for parallel processing, allowing large datasets to be processed on high-performance computing systems. We demonstrate its use on several experimental datasets to show improved image quality through parameter refinement.

36 MATERIALS SCIENCE↗

Strain-Driven Stacking Faults in CdSe/CdS Core/Shell Nanorods

Colloidal semiconductor nanocrystals are commonly grown with a shell of a second semiconductor material to obtain desired physical properties, such as increased photoluminescence quantum yield. However, the growth of a lattice-mismatched shell results in strain within the nanocrystal, and this strain has the potential to produce crystalline defects. Here in this paper, we study CdSe/CdS core/shell nanorods as a model system to investigate the influence of core size and shape on the formation of stacking faults in the nanocrystal. Using a combination of high-angle annular dark-field scanning transmission electron microscopy and pair-distribution-function analysis of synchrotron X-ray scattering, we show that growth of the CdS shell on smaller, spherical CdSe cores results in relatively small strain and few stacking faults. By contrast, growth of the shell on larger, prolate spheroidal cores leads to significant strain in the CdS lattice, resulting in a high density of stacking faults.

36 MATERIALS SCIENCE↗

Stochasticity of Pores Interconnectivity in Li–O 2 Batteries and its Impact on the Variations in Electrochemical Performance

While large dispersions in electrochemical performance have been reported for lithium oxygen batteries in the literature, they have not been investigated in any depth. The variability in the results is often assumed to arise from differences in cell design, electrode structure, handling and cell preparation at different times. An accurate theoretical framework turns out to be needed to get a better insight into the mechanisms underneath and to interpreting experimental results. Here, we develop and use a pore network model to simulate the electrochemical performance of three-dimensionally resolved lithium-oxygen cathode mesostructures obtained from TXM nano-computed tomography images. Here, we apply this model to the 3D reconstructed object of a Super P carbon electrode and calculate discharge curves, using identical conditions, for four different zones in the electrode and their reversed configurations. The resulting galvanostatic discharge curves show some dispersion, (both in terms of capacity and overpotential) which we attribute to the way pores are connected with each other. Based on these results, we propose that the stochastic nature of pores interconnectivity and the microscopic arrangement of pores can lead, at least partially, to the variations in electrochemical results observed experimentally.

25 ENERGY STORAGE↗

Reversible magnesium and aluminium ions insertion in cation-deficient anatase TiO 2

In contrast to monovalent lithium or sodium ions, the reversible insertion of multivalent ions such as Mg 2+ and Al 3+ into electrode materials remains an elusive goal. In this work, we demonstrate a new strategy to achieve reversible Mg 2+ and Al 3+ insertion in anatase TiO 2 , achieved through aliovalent doping, to introduce a large number of titanium vacancies that act as intercalation sites. We present a broad range of experimental and theoretical characterizations that show a preferential insertion of multivalent ions into titanium vacancies, allowing a much greater capacity to be obtained compared to pure TiO 2 . In conclusion, this result highlights the possibility to use the chemistry of defects to unlock the electrochemical activity of known materials providing a new strategy for the chemical design of materials for practical multivalent batteries.

25 ENERGY STORAGE↗

Self-assembled tunable networks of sticky colloidal particles

Self-assembled tunable networks of microscopic polymer fibers ranging from wavy colloidal "fur" to highly interconnected networks are created from polymer systems and an applied electric field. The networks emerge via dynamic self-assembly in an alternating (ac) electric field from a non-aqueous suspension of "sticky" polymeric colloidal particles with a controlled degree of polymerization. The resulting architectures are tuned by the frequency and amplitude of the electric field and surface properties of the particles.

36 MATERIALS SCIENCE↗