A5075010416- Machine Learning in Materials Science
- Advanced Battery Materials and Technologies
- Advanced Thermoelectric Materials and Devices
- Advancements in Battery Materials
- Ionic liquids properties and applications
- Protein Structure and Dynamics
- Thermal properties of materials
- Catalytic Processes in Materials Science
- Fuel Cells and Related Materials
- Catalysis and Oxidation Reactions
- nanoparticles nucleation surface interactions
- Electrochemical Analysis and Applications
- Computational Drug Discovery Methods
- Rare-earth and actinide compounds
- Electronic and Structural Properties of Oxides
- Advanced Materials Characterization Techniques
- Advanced Chemical Physics Studies
- Heusler alloys: electronic and magnetic properties
- Conducting polymers and applications
- Graphene research and applications
- Advanced Battery Technologies Research
- Mass Spectrometry Techniques and Applications
- Topic Modeling
- Ferroelectric and Piezoelectric Materials
- X-ray Diffraction in Crystallography
Robert Bosch (United States)
2015-2025
Harvard University
2018-2025
Harvard University Press
2018-2025
Robert Bosch (Slovenia)
2011-2022
Ewha Womans University
2019
University of Geneva
2018
National Institute of Standards and Technology
2018
New York University Press
2018
MicroVision (United States)
2018
Pennsylvania State University
2018
Abstract This work presents Neural Equivariant Interatomic Potentials (NequIP), an E(3)-equivariant neural network approach for learning interatomic potentials from ab-initio calculations molecular dynamics simulations. While most contemporary symmetry-aware models use invariant convolutions and only act on scalars, NequIP employs interactions of geometric tensors, resulting in a more information-rich faithful representation atomic environments. The method achieves state-of-the-art accuracy...
The thermal conductivity of disordered silicon-germanium alloys is computed from density-functional perturbation theory and with relaxation times that include both harmonic anharmonic scattering terms. We show this approach yields an excellent agreement at all compositions experimental results provides clear design rules for the engineering nanostructured thermoelectrics. For Si(x)Ge(1-x), more than 50% heat carried room temperature by phonons mean free path greater 1 μm, addition as little...
We derive octupole-level secular perturbation equations for hierarchical triple systems, using classical Hamiltonian techniques. Our describe the evolution of orbital eccentricities and inclinations over timescales that are long compared to periods. By extending previous work done leading (quadrupole) order octupole level (i.e., including terms α3, where α ≡ a1/a2 < 1 is ratio semimajor axes), we obtain expressions applicable a much wider range parameters. In particular, our results can be...
Abstract Machine learned force fields typically require manual construction of training sets consisting thousands first principles calculations, which can result in low efficiency and unpredictable errors when applied to structures not represented the set model. This severely limits practical application these models systems with dynamics governed by important rare events, such as chemical reactions diffusion. We present an adaptive Bayesian inference method for automating interpretable,...
A simultaneously accurate and computationally efficient parametrization of the potential energy surface molecules materials is a long-standing goal in natural sciences. While atom-centered message passing neural networks (MPNNs) have shown remarkable accuracy, their information propagation has limited accessible length-scales. Local methods, conversely, scale to large simulations but suffered from inferior accuracy. This work introduces Allegro, strictly local equivariant deep network...
The Li/oxygen battery may achieve a high practical specific energy as its theoretical is Li assuming the product. To help understand physics of we present first physics-based model that incorporates major thermodynamic, transport, and kinetic processes. We obtain good match between porous-electrode experiments simulations by using an empirical fit to resistance discharge products (which include carbonates oxides when carbonate solvents) function thickness obtained from flat-electrode...
Abstract The ever-growing availability of computing power and the sustained development advanced computational methods have contributed much to recent scientific progress. These developments present new challenges driven by sheer amount calculations data manage. Next-generation exascale supercomputers will harden these challenges, such that automated scalable solutions become crucial. In years, we been developing AiiDA (aiida.net), a robust open-source high-throughput infrastructure...
Currently available solid polymer electrolytes for Li-ion cells require deeper understanding and significant improvement in ionic transport properties to enable their use high-power batteries. We molecular dynamics simulations model the amorphous electrolyte system comprising poly(ethylene) oxide (PEO), lithium, bis(trifluoromethane)sulfonimide anion (TFSI), exploring effects of high salt concentrations relevant battery applications. Using statistical analysis ion distribution transport, we...
We report a peak dimensionless figure-of-merit (ZT) of ∼1 at 700 °C in nanostructured p-type Nb0.6Ti0.4FeSb0.95Sn0.05 composition. Even though the power factor composition is improved by 25%, comparison to previously reported Hf0.44Zr0.44Ti0.12CoSb0.8Sn0.2, ZT value not increased due higher thermal conductivity. However, led 15% increase output thermoelectric device made from previous best material Hf0.44Zr0.44Ti0.12CoSb0.8Sn0.2. The n-type used make unicouple Hf0.25Zr0.75NiSn0.99Sb0.01 with...
Understanding the ionic diffusion mechanism in polymer electrolytes is critical to development of advanced lithium-ion batteries. We report here molecular dynamics-based characterization structures and poly(ethylene oxide) (PEO) with lithium bis(trifluoromethysulfonyl)imide (TFSI) ions imbedded into PEO structure. consider a range temperatures (360–480 K), weights (43, 22, 10, 2 chains 23, 45, 100, 450 EO monomers, respectively), ion concentrations (r = 0.02, 0.04, 0.06, 0.08 Li:EO) for...
Abstract Recently, machine learning (ML) has been used to address the computational cost that limiting ab initio molecular dynamics (AIMD). Here, we present GNNFF, a graph neural network framework directly predict atomic forces from automatically extracted features of local environment are translationally-invariant, but rotationally-covariant coordinate atoms. We demonstrate GNNFF not only achieves high performance in terms force prediction accuracy and speed on various materials systems,...
Accurate modeling of chemically reactive systems has traditionally relied on either expensive ab initio approaches or flexible bond-order force fields such as ReaxFF that require considerable time, effort, and expertise to parameterize. Here, we introduce FLARE++, a Bayesian active learning method for training many-body the fly during molecular dynamics (MD) simulations. During automated loop, predictive uncertainties sparse Gaussian process (SGP) field are evaluated at each timestep an MD...
Abstract This work examines challenges associated with the accuracy of machine-learned force fields (MLFFs) for bulk solid and liquid phases d -block elements. In exhaustive detail, we contrast performance force, energy, stress predictions across transition metals two leading MLFF models: a kernel-based atomic cluster expansion method implemented using sparse Gaussian processes (FLARE), an equivariant message-passing neural network (NequIP). Early present higher relative errors are more...
Abstract Molecular dynamics simulation is an important tool in computational materials science and chemistry, the past decade it has been revolutionized by machine learning. This rapid progress learning interatomic potentials produced a number of new architectures just few years. Particularly notable among these are atomic cluster expansion, which unified many earlier ideas around atom-density-based descriptors, Neural Equivariant Interatomic Potentials (NequIP), message-passing neural...
We characterize the response of isolated single-wall (SWNT) and multiwall (MWNT) carbon nanotubes nanotube bundles to static electric fields using first-principles calculations density-functional theory. The longitudinal polarizability SWNTs scales as inverse square band gap, while in MWNTs it is given by sum polarizabilities constituent tubes. transverse insensitive gaps chiralities proportional effective radius; MWNTs, outer layers dominate response. intermediate between metallic...
We address periodic-image errors arising from the use of periodic boundary conditions to describe systems that do not exhibit full three-dimensional periodicity. The difference between potential, as straightforwardly obtained a Fourier transform, and potential satisfying any other can be characterized analytically. In light this observation, we present an efficient real-space method correct errors, based on multigrid solver for difference, demonstrate excellent convergence energy with...
We present a large-scale density functional theory (DFT) investigation of the $\mathit{AB}{\mathrm{O}}_{3}$ chemical space in perovskite crystal structure, with aim identifying those that are relevant for forming piezoelectric materials. Screening criteria on DFT results used to select 49 compositions, which can be seen as fundamental building blocks from create alloys potentially good performance. This screening finds all alloy end points three well-known high-performance piezoelectrics....
We show that strong cation-anion interactions in a wide range of lithium-salt/ionic liquid mixtures result negative lithium transference number, using molecular dynamics simulations and rigorous concentrated solution theory. This behavior fundamentally deviates from obtained self-diffusion coefficient analysis explains well recent experimental electrophoretic nuclear magnetic resonance measurements, which account for ion correlations. extend these findings to several ionic compositions....
Quasi-elastic neutron scattering experiments on mixtures of poly(ethylene oxide) and lithium bis(trifluoromethane)sulfonimide salt, a standard polymer electrolyte, led to the quantification effect salt segmental dynamics in 1-10 Å length scale. The monomeric friction coefficient characterizing these scales increases exponentially with concentration. More importantly, we find that this change alone is responsible for all observed nonlinearity dependence ionic conductivity Our analysis leads...
The Open Databases Integration for Materials Design (OPTIMADE) consortium has designed a universal application programming interface (API) to make materials databases accessible and interoperable. We outline the first stable release of specification, v1.0, which is already supported by many leading several software packages. illustrate advantages OPTIMADE API through worked examples on each public that support full specification.