TY - CONF A1 - Ertural, Christina T1 - Automation of machine learning driven interatomic potential generation for predicting vibrational properties N2 - Investigating the phononic properties is beneficial for predicting low thermal conductivity thermoelectric materials.1–3 Employing density functional theory4 takes many calculation steps and consumes a lot of computational resources.5,6 Using machine learning driven interatomic potentials (MLIP, e.g., Gaussian approximation potential,8 GAP) opens up a faster route to phonons7 but in most cases, the potentials are specifically tailored for a certain compound. In this work, we automate the generation of such MLIPs in a Python code-based workflow, based on the automation tools atomate29 and pymatgen10 which combines the automatic DFT computations with the automated fitting of GAPs. Automation enables easier testing, benchmarking and validation.11 We aim to provide the workflow-generated potentials for storage in databases. T2 - 16th International conference on materials chemistry (MC16) CY - Dublin, Ireland DA - 03.07.2023 KW - Interatomic potentials KW - Machine learning KW - Phonons KW - Thermoelectrics PY - 2023 AN - OPUS4-57932 LA - eng AD - Bundesanstalt fuer Materialforschung und -pruefung (BAM), Berlin, Germany ER - TY - CONF A1 - Bustamante, Joana A1 - Naik, Aakash A1 - Ertural, Christina A1 - Ueltzen, Katharina A1 - George, Janine T1 - Thermodynamic and Thermoelectric Properties of the Canfieldite, (Ag8SnS6 ), in the Quasi-Harmonic Approximation N2 - Argyrodite-type materials have lately sparked a lot of research interest due to their thermoelectric properties.One promising candidate is canfieldite (Ag8SnS6), which has a Pna21 orthorhombic crystal structure at room temperature (RT). Recently, a new low-temperature (LT) phase transition of canfieldite at 120K has been found. Here, we investigate structural, vibrational and thermodynamic properties of Ag8SnS6 at room- and low-temperature employing density functional theory (DFT) and lattice dynamics computations. Thermal properties calculations were based on the quasi-harmonic approximation (QHA) as implemented in phonopy. We achieve good agreement with experiments. Thermal properties such as the constant-pressure heat capacity (Cp) and thermal conductivity are very close to experimental measurements. Furthermore, we compared RT and LT Ag8SnS6 Grüneisen parameters with argyrodites analogues, Ag8XS6 (X = Sn, Si, Ge), to arrive at an improved T2 - 16th International conference on materials chemistry (MC16) CY - Dublin, Ireland DA - 03.07.2023 KW - Thermoelectric materials KW - DFT KW - QHA KW - Grüneisen parameter PY - 2023 AN - OPUS4-57887 LA - eng AD - Bundesanstalt fuer Materialforschung und -pruefung (BAM), Berlin, Germany ER - TY - CONF A1 - Amariamir, Sasan A1 - Benner, Philipp A1 - George, Janine T1 - Prediction of materials synthesizability N2 - In the pursuit of discovering materials with desirable properties, extending the available material libraries is crucial. High-throughput simulations have become an integral part in designing new materials in the past decades. However, there is no straightforward way of distinguishing synthesizable materials from all the proposed candidates. This project focuses on employing AI-driven methods to estimate synthesizability of materials. Up to now, material scientists and engineers have relied on domain knowledge as well as empirical heuristics to guess the stability and synthesizability of molecules and crystals. The famous Pauling rules of crystal stability are an example of such heuristics. However, after the accelerating material discovery in all the years since Pauling, these rules now fail to account for the stability of most known crystals. A new predictive set of heuristics for crystal stability/synthesizability is unlikely to be uncovered by human perception, given the magnitude and dimensionality of crystallographic data. Hence, a data-driven approach should be proposed to find a predictive model or set of heuristics which differentiate synthesizable crystal structures from the rest. The main challenge of this research problem is the lack of a negative set for classification. Here, there are two classes of data: the positive class which contains synthesizable materials and the negative class which contains materials which are not synthesizable. While the data from the positive class is simply the data of crystals which have been experimentally synthesized, we do not have access to data points which are certainly unsynthesizable. Strictly speaking, if an attempt of synthesizing a crystal fails, it does not necessarily follow that the crystal is not synthesizable. Also, there is no database available which contains the intended crystal structures of unsuccessful synthesis attempts. This project proposes a semi-supervised learning scheme to predict crystal synthesizability. The ML model is trained on experimental and theoretical crystal data. The initial featurization focuses on local environments which is inspired by the Pauling Rules. The experimental data points are downloaded through the Pymatgen API from the Materials Project database which contains relaxed structures recorded in Inorganic Crystal Structure Database – ICSD. The theoretical data is queried from select databases accessible through the Optimade project’s API. T2 - MSE Congress 2022 CY - Darmstadt, Germany DA - 27.09.2022 KW - Synthesizability KW - PU Learning KW - Cheminformatics PY - 2022 AN - OPUS4-56731 LA - eng AD - Bundesanstalt fuer Materialforschung und -pruefung (BAM), Berlin, Germany ER - TY - CONF A1 - Waske, Anja A1 - Fähler, Sebastian T1 - Thermomagnetic generators with magnetocaloric materials for harvesting low grade waste heat N2 - To date, there are very few technologies available for the conversion of low-temperature waste heat into electricity. Thermomagnetic generators are one approach proposed more than a century ago. Such devices are based on a cyclic change of magnetization with temperature. This switches a magnetic flux and, according to Faraday’s law, induces a voltage. Here we give an overview on our research, covering both materials and systems. We demonstrate that guiding the magnetic flux with an appropriate topology of the magnetic circuit improves the performance of thermomagnetic generators by orders of magnitude. Through a combination of experiments and simulations, we show that a pretzel-like topology results in a sign reversal of the magnetic flux. This avoids the drawbacks of previous designs, namely, magnetic stray fields, hysteresis and complex geometries of the thermomagnetic material. Though magnetocaloric materials had been the first choice also for thermomagnetic generators, they require some different properties, which we illustrate with Ashby plots for materials selection. Experimentally we compare La-Fe-Co-Si and Gd plates in the same thermomagnetic generator. Furthermore, we discuss corrosion and deterioration under cyclic use is a severe problem occurring during operation. To amend this, composite plates using polymer as a matrix have been suggested previously. T2 - Dresden Days of Magnetocalorics CY - Dresden, Germany DA - 13.11.2023 KW - Thermomagnetic material KW - Waste heat recovery KW - Generator PY - 2023 AN - OPUS4-58865 LA - eng AD - Bundesanstalt fuer Materialforschung und -pruefung (BAM), Berlin, Germany ER - TY - CONF A1 - Gupta, Kanhaiya T1 - Classification of pores from micro-XCT data of additively manufactured components prepared by laser powder bed fusion. N2 - Laser powder bed fusion (PBF-LB/M) is a metal-based additive manufacturing (AM) process that utilizes a laser to melt powder particles selectively and form the desired product through a layer-by-layer fabrication process. It allows access to a vast new design space and can produce parts with far greater geometrical complexity than traditional methods. However, PBF-LB/M induces inevitable microstructural defects, such as voids and pores, that can adversely affect the quality and performance of the manufactured components. Hence, knowledge of pore formation, types, and suppression is essential for successful future AM applications. This research investigates the formation of different types of pores and their quantification from X-ray micro-computed tomography images of a formerly reported sample. It also examines the influence of the volumetric energy density (VED) on the size, shape, and location of pores for 316L stainless steel parts produced by the PBF-LB/M process. T2 - FEMS EUROMAT 2023 CY - Frankfurt a. M., Germany DA - 03.09.2023 KW - Additive manufacturing KW - Laser Powder Bed Fusion KW - X-ray Computed Tomography PY - 2023 AN - OPUS4-58337 LA - eng AD - Bundesanstalt fuer Materialforschung und -pruefung (BAM), Berlin, Germany ER - TY - CONF A1 - Waske, Anja T1 - A unique authenticator for additively manufactured parts derived from 3D microstructural information N2 - Additive manufacturing (AM) is rapidly emerging from rapid prototyping to industrial production [1]. Thus, providing AM parts with a tagging feature that allows identification, like a fingerprint, can be crucial for logistics, certification, and anti-counterfeiting purposes since nearly any geometry can be produced by AM with stolen data or reverse engineering of an original product. However, the mechanical and functional properties of the replicated part may not be identical to the original ones and pose a safety risk [2]. Several methods are already available, which range from encasing a detector to leveraging the stochastic defects of AM parts for the identification, authentication, and traceability of AM components. The most prevailing solution consists of local process manipulation, such as printing a quick response (QR) code [3] or a set of blind holes on the surface of the internal cavity of hollow components. Local manipulation of components may alter the properties. The external tagging features can be altered or even removed by post-processing treatments. Integrating electronic systems [4] in AM parts can be used to identify and authenticate components with complex or customized geometries. However, metal-based AM, especially in powder bed fusion (PBF-LB/M) techniques, has a strong shielding effect that interferes with the communication between the reader and the transponder. Our work suggests a methodology for the identification, authentication, and traceability of AM components using microstructural features in AM components. We will show a workflow that includes analysing 3D micro computed tomography data and selecting a set number of voids that fulfil the identification criteria. We will show the results this workflow produces for a series of 20 Al-based cuboid samples with identical processing parameters and discuss their prospects and limitations. The workflow can help to establish a non-tamperable connection between an additively manufactured part and its digital data and hence link the physical and the digital world. T2 - MSE Konferenz CY - Darmstadt, Germany DA - 24.09.2024 KW - Additive Manufacturing KW - Fingerprint KW - Computed tomography PY - 2024 AN - OPUS4-62288 LA - eng AD - Bundesanstalt fuer Materialforschung und -pruefung (BAM), Berlin, Germany ER - TY - CONF A1 - George, Janine T1 - Automated bonding analysis based on crystal orbital Hamilton populations N2 - We created a workflow that fully automates bonding analysis using Crystal Orbital Hamilton Populations, which are bond-weighted densities of states. This enables understanding of crystalline material properties based on chemical bonding information. To facilitate data analysis and machine-learning research, our tools include automatic plots, automated text output, and output in machine-readable format. T2 - Sommersymposium des Fördervereins Chemieolympiade CY - Online meeting DA - 25.06.2022 KW - Bonding Analysis KW - Automation KW - DFT PY - 2022 AN - OPUS4-55409 LA - eng AD - Bundesanstalt fuer Materialforschung und -pruefung (BAM), Berlin, Germany ER - TY - CONF A1 - Ertural, Christina T1 - Vibe Check via Machine Learning: Testing the Prototype N2 - Vibrational properties play a key role in determining the stability and thermal conductivity behaviour of materials. The quasi- harmonic approximation gives insight into the phononic properties of a compound, but in the established way, i.e. density functional theory based methods, it takes many calculation steps and consumes a lot of resources to arrive at the desired results. Machine learning (ML) trained interatomic potentials (e.g. Gaussian approximation potential, GAP) pose an alternative to the traditional computation way of phonons. We develop a Python code based workflow which combines automation tools like atomate2 with ML to ease providing interactomic potentials for (quantum chemical) computations and databases. T2 - #RSCPoster Twitter conference 2023 CY - Online meeting DA - 28.02.2023 KW - Interatomic potentials KW - Machine learning KW - Phonons KW - Workflow PY - 2023 UR - https://twitter.com/cer5814012/status/1630547004462858240 AN - OPUS4-57059 LA - eng AD - Bundesanstalt fuer Materialforschung und -pruefung (BAM), Berlin, Germany ER - TY - CONF A1 - Naik, Aakash T1 - New descriptors for materials properties based on bonding indicators N2 - Includes a summary of the Ph.D. project that deals with generating a database populated with materials bonding properties and how we intend to gain deeper insights into material properties through this research. T2 - SALSA 2022 CY - Berlin, Germany DA - 15.09.2022 KW - Bonding analysis KW - Automation KW - Materials Informatics KW - Computational Chemistry PY - 2022 AN - OPUS4-56142 LA - eng AD - Bundesanstalt fuer Materialforschung und -pruefung (BAM), Berlin, Germany ER - TY - CONF A1 - George, Janine T1 - A Quantum-Chemical Bonding Database for Solid-State Materials N2 - A deep insight into the chemistry and nature of individual chemical bonds is essential for understanding materials. Bonding analysis is expected to provide important features for large-scale data analysis and machine learning of material properties. Such information on chemical bonds can be calculated using the LOBSTER (www.cohp.de) software package, which post-processes data from modern density functional theory computations by projecting plane wave-based wave functions onto a local atomic orbital basis. We have performed bonding analysis on 1520 compounds (insulators and semiconductors) using a fully automated workflow combining the VASP and LOBSTER software packages. We then automatically evaluated the data with LobsterPy (https://github.com/jageo/lobsterpy) and provide results as a database. The projected densities of states and bonding indicators are benchmarked on VASP projections and available heuristics, respectively. Lastly, we illustrate the predictive power of bonding descriptors by constructing a machine-learning model for phononic properties, which shows an increase in prediction accuracies by 27 % (mean absolute errors) compared to a benchmark model differing only by not relying on any quantum-chemical bonding features. T2 - Sommersymposium des Fördervereins Chemieolympiade CY - Online meeting DA - 15.04.2023 KW - Automation KW - Chemical Bonds KW - DFT KW - Quantum Chemistry PY - 2023 AN - OPUS4-57310 LA - eng AD - Bundesanstalt fuer Materialforschung und -pruefung (BAM), Berlin, Germany ER -