Adaptively driven X-ray diffraction guided by machine learning for autonomous phase identification

Machine learning(ML)has become a valuable tool to assist and improve materials characterization,enabling automated interpretation of experimental results with techniques such as X-ray diffraction(XRD)and electron microscopy.Because ML models are fast once trained,there is a key opportunity to bring interpretation in-line with experiments and make on-the-fly decisions to achieve optimal measurement effectiveness,which creates broad opportunities for rapid learning and information extraction from experiments.Here,we demonstrate such a capability with the development of autonomous and adaptive XRD.By coupling an ML algorithm with a physical diffractometer,this method integrates diffraction and analysis such that early experimental information is leveraged to steer measurements toward features that improve the confidence of a model trained to identify crystalline phases.We validate the effectiveness of an adaptive approach by showing that ML-driven XRD can accurately detect trace amounts of materials in multi-phase mixtures with short measurement times.The improved speed of phase detection also enables in situ identification of short-lived intermediate phases formed during solid-state reactions using a standard in-house diffractometer.Our findings showcase the advantages of in-line ML for materials characterization and point to the possibility of more general approaches for adaptive experimentation.

Effective mass and Fermi surface complexity factor from ab initio band structure calculations

The effective mass is a convenient descriptor of the electronic band structure used to characterize the density of states and electron transport based on a free electron model.While effective mass is an excellent first-order descriptor in real systems,the exact value can have several definitions,each of which describe a different aspect of electron transport.Here we use Boltzmann transport calculations applied to ab initio band structures to extract a density-of-states effective mass from the Seebeck Coefficient and an inertial mass from the electrical conductivity to characterize the band structure irrespective of the exact scattering mechanism.We identify a Fermi Surface Complexity Factor:N_(v)^(*)K^(*) from the ratio of these two masses,which in simple cases depends on the number of Fermi surface pockets eN_(v)^(*) T and their anisotropy K^(*),both of which are beneficial to high thermoelectric performance as exemplified by the high values found in PbTe.The Fermi Surface Complexity factor can be used in high-throughput search of promising thermoelectric materials.

A critical examination of compound stability predictions from machine-learned formation energies

Machine learning has emerged as a novel tool for the efficient prediction of material properties,and claims have been made that machine-learned models for the formation energy of compounds can approach the accuracy of Density Functional Theory(DFT).The models tested in this work include five recently published compositional models,a baseline model using stoichiometry alone,and a structural model.By testing seven machine learning models for formation energy on stability predictions using the Materials Project database of DFT calculations for 85,014 unique chemical compositions,we show that while formation energies can indeed be predicted well,all compositional models perform poorly on predicting the stability of compounds,making them considerably less useful than DFT for the discovery and design of new solids.Most critically,in sparse chemical spaces where few stoichiometries have stable compounds,only the structural model is capable of efficiently detecting which materials are stable.The nonincremental improvement of structural models compared with compositional models is noteworthy and encourages the use of structural models for materials discovery,with the constraint that for any new composition,the ground-state structure is not known a priori.This work demonstrates that accurate predictions of formation energy do not imply accurate predictions of stability,emphasizing the importance of assessing model performance on stability predictions,for which we provide a set of publicly available tests.

Efficient first-principles prediction of solid stability:Towards chemical accuracy

The question of material stability is of fundamental importance to any analysis of system properties in condensed matter physics and materials science.The ability to evaluate chemical stability,i.e.,whether a stoichiometry will persist in some chemical environment,and structure selection,i.e.what crystal structure a stoichiometry will adopt,is critical to the prediction of materials synthesis,reactivity and properties.Here,we demonstrate that density functional theory,with the recently developed strongly constrained and appropriately normed(SCAN)functional,has advanced to a point where both facets of the stability problem can be reliably and efficiently predicted for main group compounds,while transition metal compounds are improved but remain a challenge.SCAN therefore offers a robust model for a significant portion of the periodic table,presenting an opportunity for the development of novel materials and the study of fine phase transformations even in largely unexplored systems with little to no experimental data.

Computational understanding of Li-ion batteries

Correction to:npj Computational Materials(2016)2,16002;doi:10.1038/npjcompumats.2016.2;published online 18 March 2016 Since the online publication of the above article,it has been noted that an acknowledgement section should have been included and the text should read:‘This work was supported primarily by the U.S.Department of Energy(DOE)under Contract No.DE-FG02-96ER45571.’.

Semi-supervised machine-learning classification of materials synthesis procedures

Digitizing large collections of scientific literature can enable new informatics approaches for scientific analysis and meta-analysis.However,most content in the scientific literature is locked-up in written natural language,which is difficult to parse into databases using explicitly hard-coded classification rules.In this work,we demonstrate a semi-supervised machine-learning method to classify inorganic materials synthesis procedures from written natural language.Without any human input,latent Dirichlet allocation can cluster keywords into topics corresponding to specific experimental materials synthesis steps,such as“grinding”and“heating”,“dissolving”and“centrifuging”,etc.Guided by a modest amount of annotation,a random forest classifier can then associate these steps with different categories of materials synthesis,such as solid-state or hydrothermal synthesis.Finally,we show that a Markov chain representation of the order of experimental steps accurately reconstructs a flowchart of possible synthesis procedures.Our machine-learning approach enables a scalable approach to unlock the large amount of inorganic materials synthesis information from the literature and to process it into a standardized,machine-readable database.

Construction of ground-state preserving sparse lattice models for predictive materials simulations

First-principles based cluster expansion models are the dominant approach in ab initio thermodynamics of crystalline mixtures enabling the prediction of phase diagrams and novel ground states.However,despite recent advances,the construction of accurate models still requires a careful and time-consuming manual parameter tuning process for ground-state preservation,since this property is not guaranteed by default.In this paper,we present a systematic and mathematically sound method to obtain cluster expansion models that are guaranteed to preserve the ground states of their reference data.The method builds on the recently introduced compressive sensing paradigm for cluster expansion and employs quadratic programming to impose constraints on the model parameters.The robustness of our methodology is illustrated for two lithium transition metal oxides with relevance for Li-ion battery cathodes,i.e.,Li_(2x)Fe_(2(1−x))O_(2) and Li_(2x)Ti_(2(1−x))O_(2),for which the construction of cluster expansion models with compressive sensing alone has proven to be challenging.We demonstrate that our method not only guarantees ground-state preservation on the set of reference structures used for the model construction,but also show that out-of-sample ground-state preservation up to relatively large supercell size is achievable through a rapidly converging iterative refinement.This method provides a general tool for building robust,compressed and constrained physical models with predictive power.

Computational understanding of Li-ion batteries This article has been corrected since publication and a corrigendum has also been published

Over the last two decades,computational methods have made tremendous advances,and today many key properties of lithium-ion batteries can be accurately predicted by first principles calculations.For this reason,computations have become a cornerstone of battery-related research by providing insight into fundamental processes that are not otherwise accessible,such as ionic diffusion mechanisms and electronic structure effects,as well as a quantitative comparison with experimental results.The aim of this review is to provide an overview of state-of-the-art ab initio approaches for the modelling of battery materials.We consider techniques for the computation of equilibrium cell voltages,0-Kelvin and finite-temperature voltage profiles,ionic mobility and thermal and electrolyte stability.The strengths and weaknesses of different electronic structure methods,such as DFT+U and hybrid functionals,are discussed in the context of voltage and phase diagram predictions,and we review the merits of lattice models for the evaluation of finite-temperature thermodynamics and kinetics.With such a complete set of methods at hand,first principles calculations of ordered,crystalline solids,i.e.,of most electrode materials and solid electrolytes,have become reliable and quantitative.However,the description of molecular materials and disordered or amorphous phases remains an important challenge.We highlight recent exciting progress in this area,especially regarding the modelling of organic electrolytes and solid–electrolyte interfaces.

Approaches for handling high-dimensional cluster expansions of ionic systems

Disordered multicomponent systems attract great interest due to their engineering design flexibility and subsequent rich space of properties.However,detailed characterization of the structure and atomic correlations remains challenging and hinders full navigation of these complex spaces.A lattice cluster expansion is one tool to obtain configurational and energetic resolution.While in theory a cluster expansion can be applied to any system of any dimensionality,the method has primarily been used in binary systems or ternary alloys.Here we apply cluster expansions in high-component ionic systems,setting up the largest cluster expansion ever attempted to our knowledge.In doing so,we address and discuss challenges specific to high-component ionic systems,namely charge state assignments,structural relaxations,and rank-deficient systems.We introduce practical procedures to make the fitting and analysis of complex systems tractable,providing guidance for future computational studies of disordered ionic systems.