NMR-TS: de novo molecule identification from NMR spectra.

Nuclear magnetic resonance (NMR) spectroscopy is an effective tool for identifying molecules in a sample. Although many previously observed NMR spectra are accumulated in public databases, they cover only a tiny fraction of the chemical space, and molecule identification is typically accomplished manually based on expert knowledge. Herein, we propose NMR-TS, a machine-learning-based python library, to automatically identify a molecule from its NMR spectrum. NMR-TS discovers candidate molecules whose NMR spectra match the target spectrum by using deep learning and density functional theory (DFT)-computed spectra. As a proof-of-concept, we identify prototypical metabolites from their computed spectra. After an average 5451 DFT runs for each spectrum, six of the nine molecules are identified correctly, and proximal molecules are obtained in the other cases. This encouraging result implies that de novo molecule generation can contribute to the fully automated identification of chemical structures. NMR-TS is available at https://github.com/tsudalab/NMR-TS.

Materials informatics approach to understand aluminum alloys.

The relations between the mechanical properties, heat treatment, and compositions of elements in aluminum alloys are extracted by a materials informatics technique. In our strategy, a machine learning model is first trained by a prepared database to predict the properties of materials. The dependence of the predicted properties on explanatory variables, that is, the type of heat treatment and element composition, is searched using a Markov chain Monte Carlo method. From the dependencies, a factor to obtain the desired properties is investigated. Using targets of 5000, 6000, and 7000 series aluminum alloys, we extracted relations that are difficult to find via simple correlation analysis. Our method is also used to design an experimental plan to optimize the materials properties while promoting the understanding of target materials.

Predicting material properties by integrating high-throughput experiments, high-throughput ab-initio calculations, and machine learning.

High-throughput experiments (HTEs) have been powerful tools to obtain many materials data. However, HTEs often require expensive equipment. Although high-throughput ab-initio calculation (HTC) has the potential to make materials big data easier to collect, HTC does not represent the actual materials data obtained by HTEs in many cases. Here we propose using a combination of simple HTEs, HTC, and machine learning to predict material properties. We demonstrate that our method enables accurate and rapid prediction of the Kerr rotation mapping of an FexCoyNi1-x-y composition spread alloy. Our method has the potential to quickly predict the properties of many materials without a difficult and expensive HTE and thereby accelerate materials development.

Quantitative prediction of mixture toxicity of AgNO3 and ZnO nanoparticles on Daphnia magna.

Once metal-based engineered nanoparticles (NPs) are released into the aquatic environment, they are expected to interact with other existing co-contaminants. A knowledge gap exists as to how the interaction of NPs with other co-contaminants occurs. Here we selected ZnO NPs among various NPs, with Ag ion existing as a contaminant in the aquatic environment by Ag NPs widely used. A novel modeling strategy was demonstrated enabling quantitative and predictive evaluation of the aqueous mixture nanotoxicity. Individual and binary mixture toxicity tests of ZnO NPs and silver (as AgNO3) on Daphnia magna were conducted and compared to determine whether the presence of Ag ions affects the toxicity of ZnO NPs. Binary mixture toxicity was evaluated based on the concentration addition (CA) and independent action models. The CA dose-ratio dependent model was found to be the model of best fit for describing the pattern of mixture toxicity. The MIX I and MIX III suspensions (higher ratios of ZnO NPs to AgNO3) showed a synergism, whereas the MIX II suspension (lower ratio of ZnO NPs to AgNO3) showed an antagonism. The synergistic mixture toxicity at higher ratios of ZnO NPs to AgNO3 was caused by either the physiological or metabolic disturbance induced by the excessive ionic Zn or increased transport and accumulation in D. magna via the formation of complex of ionic Ag with ZnO NPs. Therefore, the toxicity level contributed via their aggregation and physicochemical properties and the dissolved ions played a crucial role in the mixture toxicities of the NPs.

Corrosion rate prediction and influencing factors evaluation of low-alloy steels in marine atmosphere using machine learning approach.

The empirical modeling methods are widely used in corrosion behavior analysis. But due to the limited regression ability of conventional algorithms, modeling objects are often limited to individual factors and specific environments. This study proposed a modeling method based on machine learning to simulate the marine atmospheric corrosion behavior of low-alloy steels. The correlations between material, environmental factors and corrosion rate were evaluated, and their influences on the corrosion behavior of steels were analyzed intuitively. By using the selected dominating factors as input variables, an optimized random forest model was established with a high prediction accuracy of corrosion rate (R2 values, 0.94 and 0.73 to the training set and testing set) to different low-alloy steel samples in several typical marine atmospheric environments. The results demonstrated that machine learning was efficient in corrosion behavior analysis, which usually involves a regression analysis of multiple factors.

Development of spectral decomposition based on Bayesian information criterion with estimation of confidence interval.

We develop an automatic peak fitting algorithm using the Bayesian information criterion (BIC) fitting method with confidence-interval estimation in spectral decomposition. First, spectral decomposition is carried out by adopting the Bayesian exchange Monte Carlo method for various artificial spectral data, and the confidence interval of fitting parameters is evaluated. From the results, an approximated model formula that expresses the confidence interval of parameters and the relationship between the peak-to-peak distance and the signal-to-noise ratio is derived. Next, for real spectral data, we compare the confidence interval of each peak parameter obtained using the Bayesian exchange Monte Carlo method with the confidence interval obtained from the BIC-fitting with the model selection function and the proposed approximated formula. We thus confirm that the parameter confidence intervals obtained using the two methods agree well. It is therefore possible to not only simply estimate the appropriate number of peaks by BIC-fitting but also obtain the confidence interval of fitting parameters.

Self-healing by design: universal kinetic model of strength recovery in self-healing ceramics.

We propose a new theoretical kinetic model of strength recovery by oxidation-induced self-healing of surface cracks in composites containing a healing agent (HA). The kinetics is a key parameter in the design of structural components that can self-heal the damage done in service. Based on three-dimensional (3D) observations of crack-gap filling, two crack-gap filling models, i.e., a bridging model and a tip-to-mouth filling model, are incorporated in the proposed kinetic model. These crack-gap filling models account for the microstructural features of the fracture surfaces, crack geometry, and oxidation kinetics of the healing-agent. Hence, the minimum and maximum remaining flaw sizes in the healed crack gaps are estimated for various healing temperatures, times, and oxygen partial pressure conditions. Further, the nonlinear elastic fracture mechanics suitable for small-sized remaining flaws, together with a statistical analysis of the original Weibull-type strength distribution, enables the prediction of upper and lower strength limits of the healed composites. Three sintered alumina matrix composites containing silicon carbide (SiC)-type HAs with various volume fractions and shapes, together with monolithic SiC ceramics, are considered. The strength of the healed-composite predicted by our model agrees well with the experimental values. This theoretical approach can be applied to HAs other than SiC and enables the design of self-healing ceramic components for various applications.

Stacking fault energy prediction for austenitic steels: thermodynamic modeling vs. machine learning.

Stacking fault energy (SFE) is of the most critical microstructure attribute for controlling the deformation mechanism and optimizing mechanical properties of austenitic steels, while there are no accurate and straightforward computational tools for modeling it. In this work, we applied both thermodynamic modeling and machine learning to predict the stacking fault energy (SFE) for more than 300 austenitic steels. The comparison indicates a high need of improving low-temperature CALPHAD (CALculation of PHAse Diagrams) databases and interfacial energy prediction to enhance thermodynamic model reliability. The ensembled machine learning algorithms provide a more reliable prediction compared with thermodynamic and empirical models. Based on the statistical analysis of experimental results, only Ni and Fe have a moderate monotonic influence on SFE, while many other elements exhibit a complex effect that their influence on SFE may change with the alloy composition.

Descriptors for dielectric constants of perovskite-type oxides by materials informatics with first-principles density functional theory.

Dielectric materials that can realize downsizing and higher performance in electric devices are in demand. Perovskite-type materials of the form ABO3 are potential candidates. However, because of the numerous conceivable compositions of perovskite-type oxides, finding the best composition is technically difficult. To obtain a reasonable guideline for material design, we aim to clarify the relationship between the dielectric constants and other physical and chemical properties of perovskite-type oxides using first-principles density functional theory (DFT) and partial least-squares regression analysis. The more important factors affecting the dielectric constants are predicted based on variable importance in projection (VIP) scores. The dielectric constant strongly correlates with the ionicity of the B cations and the density of states of the conduction bands of the B cations.

Data-based selection of creep constitutive models for high-Cr heat-resistant steel.

There are two types of creep constitutive equation, one with a steady-state term (steady-state type) and the other with no steady-state term (non-steady-state type). We applied the Bayesian inference framework in order to examine which type is supported by experimental creep curves for a Grade 91 (Gr.91) steel. The Bayesian free energy was significantly lower for the steady-state type under all the test conditions in the ranges of 50-90 MPa at 923 K, 90-160 MPa at 873 K and 170-240 MPa at 823 K, leading to the conclusion that the posterior probability was virtually 1.0. These findings mean that the experimental data supported the steady-state-type equation. The dependence of the evaluated steady-state creep rate on the applied stress indicates that there is a transition in the mechanism governing creep deformation around 120 MPa.

Essential structural and experimental descriptors for bulk and grain boundary conductivities of Li solid electrolytes.

We present a computational approach for identifying the important descriptors of the ionic conductivities of lithium solid electrolytes. Our approach discriminates the factors of both bulk and grain boundary conductivities, which have been rarely reported. The effects of the interrelated structural (e.g. grain size, phase), material (e.g. Li ratio), chemical (e.g. electronegativity, polarizability) and experimental (e.g. sintering temperature, synthesis method) properties on the bulk and grain boundary conductivities are investigated via machine learning. The data are trained using the bulk and grain boundary conductivities of Li solid conductors at room temperature. The important descriptors are elucidated by their feature importance and predictive performances, as determined by a nonlinear XGBoost algorithm: (i) the experimental descriptors of sintering conditions are significant for both bulk and grain boundary, (ii) the material descriptors of Li site occupancy and Li ratio are the prior descriptors for bulk, (iii) the density and unit cell volume are the prior structural descriptors while the polarizability and electronegativity are the prior chemical descriptors for grain boundary, (iv) the grain size provides physical insights such as the thermodynamic condition and should be considered for determining grain boundary conductance in solid polycrystalline ionic conductors.

Sparse modeling of chemical bonding in binary compounds.

A sparse model for quantifying energy difference between zinc-blende and rock-salt crystal structures in octet elemental and binary materials is constructed by using the linearly independent descriptor-generation method and exhaustive search, following the previous work by Ghiringhelli et al. [Phys Rev Lett. 2015;114:105503]. The obtained simplest model includes only atomic radius information of constituent atoms and its physical meaning is interpreted in relation to van Arkel-Ketelaar's triangle for classifying chemical bonding in binary compounds.

Data analytics approach for melt-pool geometries in metal additive manufacturing.

Modern data analytics was employed to understand and predict physics-based melt-pool formation by fabricating Ni alloy single tracks using powder bed fusion. An extensive database of melt-pool geometries was created, including processing parameters and material characteristics as input features. Correlation analysis provided insight for relationships between process parameters and melt-pools, and enabled the development of meaningful machine learning models via the use of highly correlated features. We successfully demonstrated that data analytics facilitates understanding of the inherent physics and reliable prediction of melt-pool geometries. This approach can serve as a basis for the melt-pool control and process optimization.

Unveiling the principle descriptor for predicting the electron inelastic mean free path based on a machine learning framework.

The TPP-2M formula is the most popular empirical formula for the estimation of the electron inelastic mean free paths (IMFPs) in solids from several simple material parameters. The TPP-2M formula, however, poorly describes several materials because it relies heavily on the traditional least-squares analysis. Herein, we propose a new framework based on machine learning to overcome the weakness. This framework allows a selection from an enormous number of combined terms (descriptors) to build a new formula that describes the electron IMFPs. The resulting framework not only provides higher average accuracy and stability but also reveals the physics meanings of several newly found descriptors. Using the identified principle descriptors, a complete physics picture of electron IMFPs is obtained, including both single and collective electron behaviors of inelastic scattering. Our findings suggest that machine learning is robust and efficient to predict the IMFP and has great potential in building a regression framework for data-driven problems. Furthermore, this method could be applicable to find empirical formula for given experimental data using a series of parameters given a priori, holds potential to find a deeper connection between experimental data and a priori parameters.

Automatic steel labeling on certain microstructural constituents with image processing and machine learning tools.

It is demonstrated that optical microscopy images of steel materials could be effectively categorized into classes on preset ferrite/pearlite-, ferrite/pearlite/bainite-, and bainite/martensite-type microstructures with image pre-processing and statistical analysis including the machine learning techniques. Though several popular classifiers were able to get the reasonable class-labeling accuracy, the random forest was virtually the best choice in terms of overall performance and usability. The present categorizing classifier could assist in choosing the appropriate pattern recognition method from our library for various steel microstructures, which we have recently reported. That is, the combination of the categorizing and pattern-recognizing methods provides a total solution for automatic quantification of a wide range of steel microstructures.

First-principles prediction of high oxygen-ion conductivity in trilanthanide gallates Ln3GaO6.

We systematically investigated trilanthanide gallates (Ln3GaO6) with the space group Cmc21 as oxygen-ion conductors using first-principles calculations. Six Ln3GaO6 (Ln = Nd, Gd, Tb, Ho, Dy, or Er) are both energetically and dynamically stable among 15 Ln3GaO6 compounds, which is consistent with previous experimental studies reporting successful syntheses of single phases. La3GaO6 and Lu3GaO6 may be metastable despite a slightly higher energy than those of competing reference states, as phonon calculations predict them to be dynamically stable. The formation and the migration barrier energies of an oxygen vacancy (V O) suggest that eight Ln3GaO6 (Ln = La, Nd, Gd, Tb, Ho, Dy, Er, or Lu) can act as oxygen-ion conductors based on V O. Ga plays a role of decreasing the distances between the oxygen sites of Ln3GaO6 compared with those of Ln2O3 so that a V O migrates easier with a reduced migration barrier energy. Larger oxygen-ion diffusivities and lower migration barrier energies of V O for the eight Ln3GaO6 are obtained for smaller atomic numbers of Ln having larger radii of Ln3+. Their oxygen-ion conductivities at 1000 K are predicted to have a similar order of magnitude to that of yttria-stabilized zirconia.

Relation extraction with weakly supervised learning based on process-structure-property-performance reciprocity.

In this study, we develop a computer-aided material design system to represent and extract knowledge related to material design from natural language texts. A machine learning model is trained on a text corpus weakly labeled by minimal annotated relationship data (~100 labeled relationships) to extract knowledge from scientific articles. The knowledge is represented by relationships between scientific concepts, such as {annealing, grain size, strength}. The extracted relationships are represented as a knowledge graph formatted according to design charts, inspired by the process-structure-property-performance (PSPP) reciprocity. The design chart provides an intuitive effect of processes on properties and prospective processes to achieve the certain desired properties. Our system semantically searches the scientific literature and provides knowledge in the form of a design chart, and we hope it contributes more efficient developments of new materials.

A general representation scheme for crystalline solids based on Voronoi-tessellation real feature values and atomic property data.

Increasing attention has been paid to materials informatics approaches that promise efficient and fast discovery and optimization of functional inorganic materials. Technical breakthrough is urgently requested to advance this field and efforts have been made in the development of materials descriptors to encode or represent characteristics of crystalline solids, such as chemical composition, crystal structure, electronic structure, etc. We propose a general representation scheme for crystalline solids that lifts restrictions on atom ordering, cell periodicity, and system cell size based on structural descriptors of directly binned Voronoi-tessellation real feature values and atomic/chemical descriptors based on the electronegativity of elements in the crystal. Comparison was made vs. radial distribution function (RDF) feature vector, in terms of predictive accuracy on density functional theory (DFT) material properties: cohesive energy (CE), density (d), electronic band gap (BG), and decomposition energy (Ed). It was confirmed that the proposed feature vector from Voronoi real value binning generally outperforms the RDF-based one for the prediction of aforementioned properties. Together with electronegativity-based features, Voronoi-tessellation features from a given crystal structure that are derived from second-nearest neighbor information contribute significantly towards prediction.

Data-driven exploration of new pressure-induced superconductivity in PbBi2Te4.

Candidate compounds for new thermoelectric and superconducting materials, which have narrow band gap and flat bands near band edges, were exhaustively searched by the high-throughput first-principles calculation from an inorganic materials database named AtomWork. We focused on PbBi2Te4 which has the similar electronic band structure and the same crystal structure with those of a pressure-induced superconductor SnBi2Se4 explored by the same data-driven approach. The PbBi2Te4 was successfully synthesized as single crystals using a melt and slow cooling method. The core level X-ray photoelectron spectroscopy analysis revealed Pb2+, Bi3+ and Te2- valence states in PbBi2Te4. The thermoelectric properties of the PbBi2Te4 sample were measured at ambient pressure and the electrical resistance was also evaluated under high pressure using a diamond anvil cell with boron-doped diamond electrodes. The resistance decreased with increasing of the pressure, and pressure-induced superconducting transitions were discovered at 2.5 K under 10 GPa. The maximum superconducting transition temperature increased up to 8.4 K at 21.7 GPa. The data-driven approach shows promising power to accelerate the discovery of new thermoelectric and superconducting materials.

Quantitative description on structure-property relationships of Li-ion battery materials for high-throughput computations.

Li-ion batteries are a key technology for addressing the global challenge of clean renewable energy and environment pollution. Their contemporary applications, for portable electronic devices, electric vehicles, and large-scale power grids, stimulate the development of high-performance battery materials with high energy density, high power, good safety, and long lifetime. High-throughput calculations provide a practical strategy to discover new battery materials and optimize currently known material performances. Most cathode materials screened by the previous high-throughput calculations cannot meet the requirement of practical applications because only capacity, voltage and volume change of bulk were considered. It is important to include more structure-property relationships, such as point defects, surface and interface, doping and metal-mixture and nanosize effects, in high-throughput calculations. In this review, we established quantitative description of structure-property relationships in Li-ion battery materials by the intrinsic bulk parameters, which can be applied in future high-throughput calculations to screen Li-ion battery materials. Based on these parameterized structure-property relationships, a possible high-throughput computational screening flow path is proposed to obtain high-performance battery materials.