Explore the vast range of titles available.

Delithiation of layered oxide electrodes triggers irreversible oxygen loss, one of the primary degradation modes in lithium-ion batteries. However, the delithiation-dependent mechanisms of oxygen loss remain poorly understood. Here we investigate the oxygen non-stoichiometry in Li1.18–xNi0.21Mn0.53Co0.08O2–δ electrodes as a function of Li content by using cycling protocols with long open-circuit voltage steps at varying states of charge. Surprisingly, we observe substantial oxygen loss even at moderate delithiation, corresponding to 2.5, 4.0 and 7.6 ml O2 per gram of Li1.18–xNi0.21Mn0.53Co0.08O2–δ after resting at upper capacity cut-offs of 135, 200 and 265 mAh g−1 for 100 h. Our observations suggest an intrinsic oxygen instability consistent with predictions of high oxygen activity at intermediate potentials versus Li/Li+. In addition, we observe a large chemical expansion coefficient with respect to oxygen non-stoichiometry, which is about three times greater than those of classical oxygen-deficient materials such as fluorite and perovskite oxides. Furthermore, our work challenges the conventional wisdom that deep delithiation is a necessary condition for oxygen loss in layered oxide electrodes and highlights the importance of calendar ageing for investigating oxygen stability.

This work presents Neural Equivariant Interatomic Potentials (NequIP), an E(3)-equivariant neural network approach for learning interatomic potentials from ab-initio calculations for molecular dynamics simulations. While most contemporary symmetry-aware models use invariant convolutions and only act on scalars, NequIP employs E(3)-equivariant convolutions for interactions of geometric tensors, resulting in a more information-rich and faithful representation of atomic environments. The method achieves state-of-the-art accuracy on a challenging and diverse set of molecules and materials while exhibiting remarkable data efficiency. NequIP outperforms existing models with up to three orders of magnitude fewer training data, challenging the widely held belief that deep neural networks require massive training sets. The high data efficiency of the method allows for the construction of accurate potentials using high-order quantum chemical level of theory as reference and enables high-fidelity molecular dynamics simulations over long time scales. An E(3)-equivariant deep learning interatomic potential is introduced for accelerating molecular dynamics simulations. The method obtains state-of-the-art accuracy, can faithfully describe dynamics of complex systems with remarkable sample efficiency.

The face-centered cubic medium-entropy alloy NiCoCr has received considerable attention for its good mechanical properties, uncertain stacking fault energy, etc, some of which have been attributed to chemical short-range order (SRO). Here, we examine the yield strength and misfit volumes of NiCoCr to determine whether SRO has measurably influenced mechanical properties. Polycrystalline strengths show no systematic trend with different processing conditions. Measured misfit volumes in NiCoCr are consistent with those in random binaries. Yield strength prediction of a random NiCoCr alloy matches well with experiments. Finally, we show that standard spin-polarized density functional theory (DFT) calculations of misfit volumes are not accurate for NiCoCr. This implies that DFT may be inaccurate for other subtle structural quantities such as atom-atom bond distance so that caution is required in drawing conclusions about NiCoCr based on DFT. These findings all lead to the conclusion that, under typical processing conditions, SRO in NiCoCr is either negligible or has no systematic measurable effect on strength. Chemical short-range order (SRO) NiCoCr has been proposed to account for its positive stacking fault energy and good mechanical properties. Here, a combination of theory and experiment shows that SRO is of negligible importance in NiCoCr processed by standard methods.

Deep learning (DL) is one of the fastest-growing topics in materials data science, with rapidly emerging applications spanning atomistic, image-based, spectral, and textual data modalities. DL allows analysis of unstructured data and automated identification of features. The recent development of large materials databases has fueled the application of DL methods in atomistic prediction in particular. In contrast, advances in image and spectral data have largely leveraged synthetic data enabled by high-quality forward models as well as by generative unsupervised DL methods. In this article, we present a high-level overview of deep learning methods followed by a detailed discussion of recent developments of deep learning in atomistic simulation, materials imaging, spectral analysis, and natural language processing. For each modality we discuss applications involving both theoretical and experimental data, typical modeling approaches with their strengths and limitations, and relevant publicly available software and datasets. We conclude the review with a discussion of recent cross-cutting work related to uncertainty quantification in this field and a brief perspective on limitations, challenges, and potential growth areas for DL methods in materials science.

Experiments have shown that graphene-supported Ni-single atom catalysts (Ni-SACs) provide a promising strategy for the electrochemical reduction of CO 2 to CO, but the nature of the Ni sites (Ni-N 2 C 2 , Ni-N 3 C 1 , Ni-N 4 ) in Ni-SACs has not been determined experimentally. Here, we apply the recently developed grand canonical potential kinetics (GCP-K) formulation of quantum mechanics to predict the kinetics as a function of applied potential (U) to determine faradic efficiency, turn over frequency, and Tafel slope for CO and H 2 production for all three sites. We predict an onset potential (at 10 mA cm −2 ) U onset  = −0.84 V (vs. RHE) for Ni-N 2 C 2 site and U onset  = −0.92 V for Ni-N 3 C 1 site in agreement with experiments, and U onset  = −1.03 V for Ni-N 4 . We predict that the highest current is for Ni-N 4 , leading to 700 mA cm −2 at U = −1.12 V. To help determine the actual sites in the experiments, we predict the XPS binding energy shift and CO vibrational frequency for each site. Single atom catalysts (SACs) are promising in electrocatalysis but challenging to characterize. Here, the authors apply a recently developed quantum mechanical grand canonical potential kinetics method to predict reaction mechanisms and rates for CO 2 reduction at different sites of graphene-supported Ni-SACs.

The segregation of solute atoms at grain boundaries (GBs) can profoundly impact the structural properties of metallic alloys, and induce effects that range from strengthening to embrittlement. And, though known to be anisotropic, there is a limited understanding of the variation of solute segregation tendencies across the full, multidimensional GB space, which is critically important in polycrystals where much of that space is represented. Here we develop a machine learning framework that can accurately predict the segregation tendency—quantified by the segregation enthalpy spectrum—of solute atoms at GB sites in polycrystals, based solely on the undecorated (pre-segregation) local atomic environment of such sites. We proceed to use the learning framework to scan across the alloy space, and build an extensive database of segregation energy spectra for more than 250 metal-based binary alloys. The resulting machine learning models and segregation database are key to unlocking the full potential of GB segregation as an alloy design tool, and enable the design of microstructures that maximize the useful impacts of segregation. Predicting segregation energies of alloy systems can be challenging even for a single grain boundary. Here the authors propose a machine-learning framework, which maps the local environments on a distribution of segregation energies, to predict segregation energies of alloy elements in polycrystalline materials.

The theory of grain boundary (the interface between crystallites, GB) structure has a long history 1 and the concept of GBs undergoing phase transformations was proposed 50 years ago 2 , 3 . The underlying assumption was that multiple stable and metastable states exist for different GB orientations 4 – 6 . The terminology ‘complexion’ was recently proposed to distinguish between interfacial states that differ in any equilibrium thermodynamic property 7 . Different types of complexion and transitions between complexions have been characterized, mostly in binary or multicomponent systems 8 – 19 . Simulations have provided insight into the phase behaviour of interfaces and shown that GB transitions can occur in many material systems 20 – 24 . However, the direct experimental observation and transformation kinetics of GBs in an elemental metal have remained elusive. Here we demonstrate atomic-scale GB phase coexistence and transformations at symmetric and asymmetric [ 11 1 ¯ ] tilt GBs in elemental copper. Atomic-resolution imaging reveals the coexistence of two different structures at Σ19b GBs (where Σ19 is the density of coincident sites and b is a GB variant), in agreement with evolutionary GB structure search and clustering analysis 21 , 25 , 26 . We also use finite-temperature molecular dynamics simulations to explore the coexistence and transformation kinetics of these GB phases. Our results demonstrate how GB phases can be kinetically trapped, enabling atomic-scale room-temperature observations. Our work paves the way for atomic-scale in situ studies of metallic GB phase transformations, which were previously detected only indirectly 9 , 15 , 27 – 29 , through their influence on abnormal grain growth, non-Arrhenius-type diffusion or liquid metal embrittlement. Atomic-resolution observations combined with simulations show that grain boundaries within elemental copper undergo temperature-induced solid-state phase transformation to different structures; grain boundary phases can also coexist and are kinetically trapped structures.

The atomic scale structure of the active sites in heterogeneous catalysts is central to their reactivity and selectivity. Therefore, understanding active site stability and evolution under different reaction conditions is key to the design of efficient and robust catalysts. Herein we describe theoretical calculations which predict that carbon monoxide can be used to stabilize different active site geometries in bimetallic alloys and then demonstrate experimentally that the same PdAu bimetallic catalyst can be transitioned between a single-atom alloy and a Pd cluster phase. Each state of the catalyst exhibits distinct selectivity for the dehydrogenation of ethanol reaction with the single-atom alloy phase exhibiting high selectivity to acetaldehyde and hydrogen versus a range of products from Pd clusters. First-principles based Monte Carlo calculations explain the origin of this active site ensemble size tuning effect, and this work serves as a demonstration of what should be a general phenomenon that enables in situ control over catalyst selectivity. Single-atom alloys are promising catalysts for a number of different reactions. Here, the authors demonstrate that carbon monoxide can be used to transition a PdAu catalyst between a single atom and a cluster phase which exhibit distinct selectivities for ethanol dehydrogenation.

Intercalation-type metal oxides are promising negative electrode materials for safe rechargeable lithium-ion batteries due to the reduced risk of Li plating at low voltages. Nevertheless, their lower energy and power density along with cycling instability remain bottlenecks for their implementation, especially for fast-charging applications. Here, we report a nanostructured rock-salt Nb 2 O 5 electrode formed through an amorphous-to-crystalline transformation during repeated electrochemical cycling with Li + . This electrode can reversibly cycle three lithiums per Nb 2 O 5 , corresponding to a capacity of 269 mAh g −1 at 20 mA g −1 , and retains a capacity of 191 mAh g −1 at a high rate of 1 A g −1 . It exhibits superb cycling stability with a capacity of 225 mAh g −1 at 200 mA g −1 for 400 cycles, and a Coulombic efficiency of 99.93%. We attribute the enhanced performance to the cubic rock-salt framework, which promotes low-energy migration paths. Our work suggests that inducing crystallization of amorphous nanomaterials through electrochemical cycling is a promising avenue for creating unconventional high-performance metal oxide electrode materials. Intercalation-type metal oxides are promising anodes for Li-ion batteries but suffer from low energy and power density together with cycling instability. A nanostructured rock-salt Nb 2 O 5 formed via amorphous-to-crystalline transformation during cycling with Li + is shown to exhibit enhanced performance.

The limits of dislocation-mediated metal plasticity are studied by using in situ computational microscopy to reduce the enormous amount of data from fully dynamic atomistic simulations into a manageable form. Probing plasticity limits of metals Fully dynamic atomistic simulations of plastic deformation in metals are so computationally demanding that materials physicists have instead developed mesoscale proxies to model dislocation dynamics. In this paper, Vasily Bulatov and colleagues take on the challenge of modelling metal plasticity at the atomic level. Such simulations require models that contain many millions of atoms (the largest simulation in this study contains 268 million atoms), and algorithms are used to process the datasets down to a volume that allows human interpretation. The authors probe ultrahigh-strain-rate deformation in body-centred-cubic tantalum, a model metal, to investigate the limits of metal plasticity. They show that at certain limiting conditions, dislocations can no longer relieve metal loading and twinning takes over. At a strain rate lower than this limit, flow stress and dislocation density achieve a steady state and a sort of metal kneading is observed. The simulations support previous proposals of the maximum dislocation density that can be reached before a metal collapses. Ordinarily, the strength and plasticity properties of a metal are defined by dislocations—line defects in the crystal lattice whose motion results in material slippage along lattice planes 1 . Dislocation dynamics models are usually used as mesoscale proxies for true atomistic dynamics, which are computationally expensive to perform routinely 2 . However, atomistic simulations accurately capture every possible mechanism of material response, resolving every “jiggle and wiggle” 3 of atomic motion, whereas dislocation dynamics models do not. Here we present fully dynamic atomistic simulations of bulk single-crystal plasticity in the body-centred-cubic metal tantalum. Our goal is to quantify the conditions under which the limits of dislocation-mediated plasticity are reached and to understand what happens to the metal beyond any such limit. In our simulations, the metal is compressed at ultrahigh strain rates along its [001] crystal axis under conditions of constant pressure, temperature and strain rate. To address the complexity of crystal plasticity processes on the length scales (85–340 nm) and timescales (1 ns–1μs) that we examine, we use recently developed methods of in situ computational microscopy 4 , 5 to recast the enormous amount of transient trajectory data generated in our simulations into a form that can be analysed by a human. Our simulations predict that, on reaching certain limiting conditions of strain, dislocations alone can no longer relieve mechanical loads; instead, another mechanism, known as deformation twinning (the sudden re-orientation of the crystal lattice 6 ), takes over as the dominant mode of dynamic response. Below this limit, the metal assumes a strain-path-independent steady state of plastic flow in which the flow stress and the dislocation density remain constant as long as the conditions of straining thereafter remain unchanged. In this distinct state, tantalum flows like a viscous fluid while retaining its crystal lattice and remaining a strong and stiff metal.