Learning molecular dynamics: predicting the dynamics of glasses by a machine learning simulator.

Many-body dynamics of atoms such as glass dynamics is generally governed by complex (and sometimes unknown) physics laws. This challenges the construction of atom dynamics simulations that both (i) capture the physics laws and (ii) run with little computation cost. Here, based on graph neural network (GNN), we introduce an observation-based graph network (OGN) framework to "bypass all physics laws" to simulate complex glass dynamics solely from their static structure. By taking the example of molecular dynamics (MD) simulations, we successfully apply the OGN to predict atom trajectories evolving up to a few hundred timesteps and ranging over different families of complex atomistic systems, which implies that the atom dynamics is largely encoded in their static structure in disordered phases and, furthermore, allows us to explore the capacity of OGN simulations that is potentially generic to many-body dynamics. Importantly, unlike traditional numerical simulations, the OGN simulations bypass the numerical constraint of small integration timestep by a multiplier of ≥5 to conserve energy and momentum until hundreds of timesteps, thus leapfrogging the execution speed of MD simulations for a modest timescale.

Designing self-assembling kinetics with differentiable statistical physics models.

The inverse problem of designing component interactions to target emergent structure is fundamental to numerous applications in biotechnology, materials science, and statistical physics. Equally important is the inverse problem of designing emergent kinetics, but this has received considerably less attention. Using recent advances in automatic differentiation, we show how kinetic pathways can be precisely designed by directly differentiating through statistical physics models, namely free energy calculations and molecular dynamics simulations. We consider two systems that are crucial to our understanding of structural self-assembly: bulk crystallization and small nanoclusters. In each case, we are able to assemble precise dynamical features. Using gradient information, we manipulate interactions among constituent particles to tune the rate at which these systems yield specific structures of interest. Moreover, we use this approach to learn nontrivial features about the high-dimensional design space, allowing us to accurately predict when multiple kinetic features can be simultaneously and independently controlled. These results provide a concrete and generalizable foundation for studying nonstructural self-assembly, including kinetic properties as well as other complex emergent properties, in a vast array of systems.

Kohn-Sham Equations as Regularizer: Building Prior Knowledge into Machine-Learned Physics.

Including prior knowledge is important for effective machine learning models in physics and is usually achieved by explicitly adding loss terms or constraints on model architectures. Prior knowledge embedded in the physics computation itself rarely draws attention. We show that solving the Kohn-Sham equations when training neural networks for the exchange-correlation functional provides an implicit regularization that greatly improves generalization. Two separations suffice for learning the entire one-dimensional H_{2} dissociation curve within chemical accuracy, including the strongly correlated region. Our models also generalize to unseen types of molecules and overcome self-interaction error.

Combining Superionic Conduction and Favorable Decomposition Products in the Crystalline Lithium-Boron-Sulfur System: A New Mechanism for Stabilizing Solid Li-Ion Electrolytes.

We report a solid-state Li-ion electrolyte predicted to exhibit simultaneously fast ionic conductivity, wide electrochemical stability, low cost, and low mass density. We report exceptional density functional theory (DFT)-based room-temperature single-crystal ionic conductivity values for two phases within the crystalline lithium-boron-sulfur (Li-B-S) system: 62 (+9, -2) mS cm-1 in Li5B7S13 and 80 (-56, -41) mS cm-1 in Li9B19S33. We report significant ionic conductivity values for two additional phases: between 0.0056 and 0.16 mS/cm -1 in Li2B2S5 and between 0.0031 and 9.7 mS cm-1 in Li3BS3 depending on the room-temperature extrapolation scheme used. To our knowledge, our prediction gives Li9B19S33 and Li5B7S13 the second and third highest reported DFT-computed single-crystal ionic conductivities of any crystalline material. We compute the thermodynamic electrochemical stability window widths of these materials to be 0.50 V for Li5B7S13, 0.16 V for Li2B2S5, 0.45 V for Li3BS3, and 0.60 V for Li9B19S33. Individually, these materials exhibit similar or better ionic conductivity and electrochemical stability than the best-known sulfide-based solid-state Li-ion electrolyte materials, including Li10GeP2S12 (LGPS). However, we predict that electrolyte materials synthesized from a range of compositions in the Li-B-S system may exhibit even wider thermodynamic electrochemical stability windows of 0.63 V and possibly as high as 3 V or greater. The Li-B-S system also has a low elemental cost of approximately 0.05 USD/m2 per 10 µm thickness, which is significantly lower than that of germanium-containing LGPS, and a comparable mass density below 2 g/cm3. These fast-conducting phases were initially brought to our attention by a machine learning-based approach to screen over 12,000 solid electrolyte candidates, and the evidence provided here represents an inspiring success for this model.

Erratum: Accelerated Search and Design of Stretchable Graphene Kirigami Using Machine Learning [Phys. Rev. Lett. 121, 255304 (2018)].

This corrects the article DOI: 10.1103/PhysRevLett.121.255304.

Screening billions of candidates for solid lithium-ion conductors: A transfer learning approach for small data.

Machine learning (ML) methods have the potential to revolutionize materials design, due to their ability to screen materials efficiently. Unlike other popular applications such as image recognition or language processing, large volumes of data are not available for materials design applications. Here, we first show that a standard learning approach using generic descriptors does not work for small data, unless it is guided by insights from physical equations. We then propose a novel method for transferring such physical insights onto more generic descriptors, allowing us to screen billions of unknown compositions for Li-ion conductivity, a scale which was previously unfeasible. This is accomplished by using the accurate model trained with physical insights to create a large database, on which we train a new ML model using the generic descriptors. Unlike previous applications of ML, this approach allows us to screen materials which have not necessarily been tested before (i.e., not on ICSD or Materials Project). Our method can be applied to any materials design application where a small amount of data is available, combined with high details of physical understanding.

Revealing the Spectrum of Unknown Layered Materials with Superhuman Predictive Abilities.

We discover the chemical composition of over 1000 materials that are likely to exhibit layered and 2D phases but have yet to be synthesized. This includes two materials our calculations indicate can exist in distinct structures with different band gaps, expanding the short list of 2D phase-change materials. Whereas databases of over 1000 layered materials have been reported, we provide the first full database of materials that are likely layered but are yet to be synthesized, providing a roadmap for the synthesis community. We accomplish this by combining physics with machine learning on experimentally obtained data and verify a subset of candidates using density functional theory. We find that our model performs five times better than practitioners in the field at identifying layered materials and is comparable to or better than professional solid-state chemists. Finally, we find that semisupervised learning can offer benefits for materials design where labels for some of the materials are unknown.

Machine learning determination of atomic dynamics at grain boundaries.

In polycrystalline materials, grain boundaries are sites of enhanced atomic motion, but the complexity of the atomic structures within a grain boundary network makes it difficult to link the structure and atomic dynamics. Here, we use a machine learning technique to establish a connection between local structure and dynamics of these materials. Following previous work on bulk glassy materials, we define a purely structural quantity (softness) that captures the propensity of an atom to rearrange. This approach correctly identifies crystalline regions, stacking faults, and twin boundaries as having low likelihood of atomic rearrangements while finding a large variability within high-energy grain boundaries. As has been found in glasses, the probability that atoms of a given softness will rearrange is nearly Arrhenius. This indicates a well-defined energy barrier as well as a well-defined prefactor for the Arrhenius form for atoms of a given softness. The decrease in the prefactor for low-softness atoms indicates that variations in entropy exhibit a dominant influence on the atomic dynamics in grain boundaries.

Metallic Metal-Organic Frameworks Predicted by the Combination of Machine Learning Methods and Ab Initio Calculations.

Emerging applications of metal-organic frameworks (MOFs) in electronic devices will benefit from the design and synthesis of intrinsically, highly electronically conductive MOFs. However, very few are known to exist. It is a challenging task to search for electronically conductive MOFs within the tens of thousands of reported MOF structures. Using a new strategy (i.e., transfer learning) of combining machine learning techniques, statistical multivoting, and ab initio calculations, we screened 2932 MOFs and identified 6 MOF crystal structures that are metallic at the level of semilocal DFT band theory: Mn2[Re6X8(CN)6]4 (X = S, Se,Te), Mn[Re3Te4(CN)3], Hg[SCN]4Co[NCS]4, and CdC4. Five of these structures have been synthesized and reported in the literature, but their electrical characterization has not been reported. Our work demonstrates the potential power of machine learning in materials science to aid in down-selecting from large numbers of potential candidates and provides the information and guidance to accelerate the discovery of novel advanced materials.

Accelerated Search and Design of Stretchable Graphene Kirigami Using Machine Learning.

Making kirigami-inspired cuts into a sheet has been shown to be an effective way of designing stretchable materials with metamorphic properties where the 2D shape can transform into complex 3D shapes. However, finding the optimal solutions is not straightforward as the number of possible cutting patterns grows exponentially with system size. Here, we report on how machine learning (ML) can be used to approximate the target properties, such as yield stress and yield strain, as a function of cutting pattern. Our approach enables the rapid discovery of kirigami designs that yield extreme stretchability as verified by molecular dynamics (MD) simulations. We find that convolutional neural networks, commonly used for classification in vision tasks, can be applied for regression to achieve an accuracy close to the precision of the MD simulations. This approach can then be used to search for optimal designs that maximize elastic stretchability with only 1000 training samples in a large design space of ∼4×10^{6} candidate designs. This example demonstrates the power and potential of ML in finding optimal kirigami designs at a fraction of iterations that would be required of a purely MD or experiment-based approach, where no prior knowledge of the governing physics is known or available.

Disconnecting structure and dynamics in glassy thin films.

Nanometrically thin glassy films depart strikingly from the behavior of their bulk counterparts. We investigate whether the dynamical differences between a bulk and thin film polymeric glass former can be understood by differences in local microscopic structure. Machine learning methods have shown that local structure can serve as the foundation for successful, predictive models of particle rearrangement dynamics in bulk systems. By contrast, in thin glassy films, we find that particles at the center of the film and those near the surface are structurally indistinguishable despite exhibiting very different dynamics. Next, we show that structure-independent processes, already present in bulk systems and demonstrably different from simple facilitated dynamics, are crucial for understanding glassy dynamics in thin films. Our analysis suggests a picture of glassy dynamics in which two dynamical processes coexist, with relative strengths that depend on the distance from an interface. One of these processes depends on local structure and is unchanged throughout most of the film, while the other is purely Arrhenius, does not depend on local structure, and is strongly enhanced near the free surface of a film.

Representations in neural network based empirical potentials.

Many structural and mechanical properties of crystals, glasses, and biological macromolecules can be modeled from the local interactions between atoms. These interactions ultimately derive from the quantum nature of electrons, which can be prohibitively expensive to simulate. Machine learning has the potential to revolutionize materials modeling due to its ability to efficiently approximate complex functions. For example, neural networks can be trained to reproduce results of density functional theory calculations at a much lower cost. However, how neural networks reach their predictions is not well understood, which has led to them being used as a "black box" tool. This lack of understanding is not desirable especially for applications of neural networks in scientific inquiry. We argue that machine learning models trained on physical systems can be used as more than just approximations since they had to "learn" physical concepts in order to reproduce the labels they were trained on. We use dimensionality reduction techniques to study in detail the representation of silicon atoms at different stages in a neural network, which provides insight into how a neural network learns to model atomic interactions.

Atomic Layer Deposition of Stable LiAlF4 Lithium Ion Conductive Interfacial Layer for Stable Cathode Cycling.

Modern lithium ion batteries are often desired to operate at a wide electrochemical window to maximize energy densities. While pushing the limit of cutoff potentials allows batteries to provide greater energy densities with enhanced specific capacities and higher voltage outputs, it raises key challenges with thermodynamic and kinetic stability in the battery. This is especially true for layered lithium transition-metal oxides, where capacities can improve but stabilities are compromised as wider electrochemical windows are applied. To overcome the above-mentioned challenges, we used atomic layer deposition to develop a LiAlF4 solid thin film with robust stability and satisfactory ion conductivity, which is superior to commonly used LiF and AlF3. With a predicted stable electrochemical window of approximately 2.0 ± 0.9 to 5.7 ± 0.7 V vs Li+/Li for LiAlF4, excellent stability was achieved for high Ni content LiNi0.8Mn0.1Co0.1O2 electrodes with LiAlF4 interfacial layer at a wide electrochemical window of 2.75-4.50 V vs Li+/Li.

Relationship between local structure and relaxation in out-of-equilibrium glassy systems.

The dynamical glass transition is typically taken to be the temperature at which a glassy liquid is no longer able to equilibrate on experimental timescales. Consequently, the physical properties of these systems just above or below the dynamical glass transition, such as viscosity, can change by many orders of magnitude over long periods of time following external perturbation. During this progress toward equilibrium, glassy systems exhibit a history dependence that has complicated their study. In previous work, we bridged the gap between structure and dynamics in glassy liquids above their dynamical glass transition temperatures by introducing a scalar field called "softness," a quantity obtained using machine-learning methods. Softness is designed to capture the hidden patterns in relative particle positions that correlate strongly with dynamical rearrangements of particle positions. Here we show that the out-of-equilibrium behavior of a model glass-forming system can be understood in terms of softness. To do this we first demonstrate that the evolution of behavior following a temperature quench is a primarily structural phenomenon: The structure changes considerably, but the relationship between structure and dynamics remains invariant. We then show that the relaxation time can be robustly computed from structure as quantified by softness, with the same relation holding both in equilibrium and as the system ages. Together, these results show that the history dependence of the relaxation time in glasses requires knowledge only of the softness in addition to the usual state variables.

Controlling O coverage and stability by alloying Au and Ag.

One of the most critical factors in oxidation catalysis is controlling the state of oxygen on the surface. Au and Ag are both effective selective oxidation catalysts for various reactions, and their interactions with oxygen are critical for determining their catalytic performance. Here, we show that the state of oxygen on a catalytic surface can be controlled by alloying Au and Ag. Using temperature programmed desorption, density functional theory (DFT), and Monte Carlo simulations, we examine how alloying Au into an Ag(110) surface affects O2 dissociation, O coverage, and O stability. DFT calculations indicate that Au resides in the second layer, in agreement with previous experimental findings. The minimum ensemble size for O2 dissociation is 2 Ag atoms in adjacent rows of the second layer. Surprisingly, adsorbed O2 and the dissociation transition state interact directly with metal atoms in the adjacent trough, such that Au in this position inhibits O2 dissociation by direct repulsion with oxygen electronic states. Using Monte Carlo simulations based on DFT energetics, we create models of the surface that agree closely with our experimental results. Both show that the O2 uptake decreases nearly linearly as the Au concentration increases, and no O2 uptake occurs for Au concentrations above 50%. For Au concentrations greater than 18%, increasing the Au concentration also decreases the stability of the adsorbed O. Based on these results, the O coverage and O stability can be tuned, in some cases independently. We also study how the reactivity of the surface is affected by these factors using CO2 oxidation as a simple test reaction.

High-Temperature Quantum Anomalous Hall Effect in n-p Codoped Topological Insulators.

The quantum anomalous Hall effect (QAHE) is a fundamental quantum transport phenomenon that manifests as a quantized transverse conductance in response to a longitudinally applied electric field in the absence of an external magnetic field, and it promises to have immense application potential in future dissipationless quantum electronics. Here, we present a novel kinetic pathway to realize the QAHE at high temperatures by n-p codoping of three-dimensional topological insulators. We provide a proof-of-principle numerical demonstration of this approach using vanadium-iodine (V-I) codoped Sb_{2}Te_{3} and demonstrate that, strikingly, even at low concentrations of ∼2% V and ∼1% I, the system exhibits a quantized Hall conductance, the telltale hallmark of QAHE, at temperatures of at least ∼50 K, which is 3 orders of magnitude higher than the typical temperatures at which it has been realized to date. The underlying physical factor enabling this dramatic improvement is tied to the largely preserved intrinsic band gap of the host system upon compensated n-p codoping. The proposed approach is conceptually general and may shed new light in experimental realization of high-temperature QAHE.

Structural Properties of Defects in Glassy Liquids.

At zero temperature a disordered solid corresponds to a local minimum in the energy landscape. As the temperature is raised or the system is driven with a mechanical load, the system explores different minima via dynamical events in which particles rearrange their relative positions. We have shown recently that the dynamics of particle rearrangements are strongly correlated with a structural quantity associated with each particle, "softness", which we can identify using supervised machine learning. Particles of a given softness have a well-defined energy scale that governs local rearrangements; because of this property, softness greatly simplifies our understanding of glassy dynamics. Here we investigate the correlation of softness with other commonly used structural quantities, such as coordination number and local potential energy. We show that although softness strongly correlates with these properties, its predictive power for rearrangement dynamics is much higher. We introduce a useful metric for quantifying the quality of structural quantities as predictors of dynamics. We hope that, in the future, authors introducing new structural measures of dynamics will compare their proposals quantitatively to softness using this metric. We also show how softness correlations give insight into rearrangements. Finally, we explore the physical meaning of softness using unsupervised dimensionality reduction and reduced curve-fitting models, and show that softness can be recast in a form that is amenable to analytical treatment.

Atomic Insight into the Lithium Storage and Diffusion Mechanism of SiO2/Al2O3 Electrodes of Lithium Ion Batteries: ReaxFF Reactive Force Field Modeling.

Atomically deposited layers of SiO2 and Al2O3 have been recognized as promising coating materials to buffer the volumetric expansion and capacity retention upon the chemo-mechanical cycling of the nanostructured silicon- (Si-) based electrodes. Furthermore, silica (SiO2) is known as a promising candidate for the anode of next-generation lithium ion batteries (LIBs) due to its superior specific charge capacity and low discharge potential similar to Si anodes. In order to describe Li-transport in mixed silica/alumina/silicon systems we developed a ReaxFF potential for Li-Si-O-Al interactions. Using this potential, a series of hybrid grand canonical Monte Carlo (GCMC) and molecular dynamic (MD) simulations were carried out to probe the lithiation behavior of silica structures. The Li transport through both crystalline and amorphous silica was evaluated using the newly optimized force field. The anisotropic diffusivity of Li in crystalline silica cases is demonstrated. The ReaxFF diffusion study also verifies the transferability of the new force field from crystalline to amorphous phases. Our simulation results indicates the capability of the developed force field to examine the energetics and kinetics of lithiation as well as Li transportation within the crystalline/amorphous silica and alumina phases and provide a fundamental understanding on the lithiation reactions involved in the Si electrodes covered by silica/alumina coating layers.

Long-Time Dynamics through Parallel Trajectory Splicing.

Simulating the atomistic evolution of materials over long time scales is a longstanding challenge, especially for complex systems where the distribution of barrier heights is very heterogeneous. Such systems are difficult to investigate using conventional long-time scale techniques, and the fact that they tend to remain trapped in small regions of configuration space for extended periods of time strongly limits the physical insights gained from short simulations. We introduce a novel simulation technique, Parallel Trajectory Splicing (ParSplice), that aims at addressing this problem through the timewise parallelization of long trajectories. The computational efficiency of ParSplice stems from a speculation strategy whereby predictions of the future evolution of the system are leveraged to increase the amount of work that can be concurrently performed at any one time, hence improving the scalability of the method. ParSplice is also able to accurately account for, and potentially reuse, a substantial fraction of the computational work invested in the simulation. We validate the method on a simple Ag surface system and demonstrate substantial increases in efficiency compared to previous methods. We then demonstrate the power of ParSplice through the study of topology changes in Ag42Cu13 core-shell nanoparticles.