Browse
Submit Data
Databases
API
Help

Dataset Information

0 Views

0 Connections

0 Citations

0 Reanalyses

0 Downloads

Omics score: 0

A machine-learned spin-lattice potential for dynamic simulations of defective magnetic iron.

ABSTRACT: A machine-learned spin-lattice interatomic potential (MSLP) for magnetic iron is developed and applied to mesoscopic scale defects. It is achieved by augmenting a spin-lattice Hamiltonian with a neural network term trained to descriptors representing a mix of local atomic configuration and magnetic environments. It reproduces the cohesive energy of BCC and FCC phases with various magnetic states. It predicts the formation energy and complex magnetic structure of point defects in quantitative agreement with density functional theory (DFT) including the reversal and quenching of magnetic moments near the core of defects. The Curie temperature is calculated through spin-lattice dynamics showing good computational stability at high temperature. The potential is applied to study magnetic fluctuations near sizable dislocation loops. The MSLP transcends current treatments using DFT and molecular dynamics, and surpasses other spin-lattice potentials that only treat near-perfect crystal cases.

SUBMITTER: Chapman JBJ

PROVIDER: S-EPMC9794737 | biostudies-literature | 2022 Dec

REPOSITORIES: biostudies-literature

ACCESS DATA

Json Xml

Publications

A machine-learned spin-lattice potential for dynamic simulations of defective magnetic iron.

Chapman Jacob B J JBJ Ma Pui-Wai PW

Scientific reports 20221227 1

A machine-learned spin-lattice interatomic potential (MSLP) for magnetic iron is developed and applied to mesoscopic scale defects. It is achieved by augmenting a spin-lattice Hamiltonian with a neural network term trained to descriptors representing a mix of local atomic configuration and magnetic environments. It reproduces the cohesive energy of BCC and FCC phases with various magnetic states. It predicts the formation energy and complex magnetic structure of point defects in quantitative agr ...[more]

PMID: 36575185

Similar Datasets

A machine-learning framework for accelerating spin-lattice relaxation simulations.

Project description:Molecular and lattice vibrations are able to couple to the spin of electrons and lead to their relaxation and decoherence. Ab initio simulations have played a fundamental role in shaping our understanding of this process but further progress is hindered by their high computational cost. Here we present an accelerated computational framework based on machine-learning models for the prediction of molecular vibrations and spin-phonon coupling coefficients. We apply this method to three open-shell coordination compounds exhibiting long relaxation times and show that this approach achieves semi-to-full quantitative agreement with ab initio methods reducing the computational cost by about 80%. Moreover, we show that this framework naturally extends to molecular dynamics simulations, paving the way to the study of spin relaxation in condensed matter beyond simple equilibrium harmonic thermal baths.

| S-EPMC11885155 | biostudies-literature

Spin-lattice-dynamics analysis of magnetic properties of iron under compression.

Project description:Compression of a magnetic material leads to a change in its magnetic properties. We examine this effect using spin-lattice dynamics for the special case of bcc-Fe, using both single- and poly-crystalline Fe and a bicontinuous nanofoam structure. We find that during the elastic phase of compression, the magnetization increases due to a higher population of the nearest-neighbor shell of atoms and the resulting higher exchange interaction of neighboring spins. In contrast, in the plastic phase of compression, the magnetization sinks, as defects are created, increasing the disorder and typically decreasing the average atom coordination number. The effects are more pronounced in single crystals than in polycrystals, since the presence of defects in the form of grain boundaries counteracts the increase in magnetization during the elastic phase of compression. Also, the effects are more pronounced at temperatures close to the Curie temperature than at room temperature. In nanofoams, the effect of compression is minor since compression proceeds more by void reduction and filament bending-with negligible effect on magnetization-than by strain within the ligaments. These findings will prove useful for tailoring magnetization under strain by introducing plasticity.

| S-EPMC10471586 | biostudies-literature

Coupled atomistic spin-lattice simulations of ultrafast demagnetization in 3d ferromagnets.

Project description:Despite decades of research, the role of the lattice and its coupling to the magnetisation during ultrafast demagnetisation processes is still not fully understood. Here we report on studies of both explicit and implicit lattice effects on laser induced ultrafast demagnetisation of bcc Fe and fcc Co. We do this using atomistic spin- and lattice dynamics simulations following a heat-conserving three-temperature model. We show that this type of Langevin-based simulation is able to reproduce observed trends of the ultrafast magnetization dynamics of fcc Co and bcc Fe. The parameters used in our models are all obtained from electronic structure theory, with the exception of the lattice dynamics damping term, where a range of parameters were investigated. It was found that while the explicit spin-lattice coupling in the studied systems does not impact the demagnetisation process notably, the lattice damping has a large influence on the details of the magnetization dynamics. The dynamics of Fe and Co following the absorption of a femtosecond laser pulse are compared with previous results for Ni and similarities and differences in the materials' behavior are analysed. For all elements investigated so far with this model, we obtain a linear relationship between the value of the maximally demagnetized state and the fluence of the laser pulse , which is in agreement with experiments. Moreover, we demonstrate that the demagnetization amplitude is largest for Ni and smallest for Co. This holds over a wide range of the reported electron-phonon couplings, and this demagnetization trend is in agreement with recent experiments.

| S-EPMC10999457 | biostudies-literature

Size-dependent Curie temperature of Ni nanoparticles from spin-lattice dynamics simulations.

Project description:The magnetic properties of Ni nanoparticles (NPs) with diameter D are investigated using spin-lattice dynamics (SLD) simulations. Using exchange interactions fitted to ab-initio results we obtain a Curie temperature, Tc , similar, but lower, than experiments. In order to reproduce quantitatively the bulk Curie temperature and the experimental results, the exchange energy has to be increased by 25% compared to the ab-initio value. During the simulated time, Ni NPs remain ferromagnetic down to the smallest sizes investigated here, containing around 500 atoms. The average magnetic moment of the NPs is slightly smaller than that determined experimentally. By considering a core-shell model for NPs, in which the shell atoms are assigned a larger magnetic moment, this discrepancy can be removed. Tc is lower for a moving lattice than for a frozen lattice, as expected, but this difference decreases with NP size because smaller NPs include higher surface disorder which dominates the transition. For NPs, Tc decreases with the NP diameter D by at most 10% at D=2 nm, in agreement with several experiments, and unlike some modeling or theoretical scaling results which predict a considerably larger decrease. The decrease of Tc is well described by finite-size scaling models, with a critical exponent that depends on the SLD settings for a frozen or moving lattice, and also depends on the procedure for determining Tc . Extrapolating the inverse of the magnetization as function of temperature near Tc gives a lower Tc than the maximum of the susceptibility.

| S-EPMC11422501 | biostudies-literature

Lattice disorder effect on magnetic ordering of iron arsenides.

Project description:This study investigates magnetic ordering temperature in nano- and mesoscale structural features in an iron arsenide. Although magnetic ground states in quantum materials can be theoretically predicted from known crystal structures and chemical compositions, the ordering temperature is harder to pinpoint due to potential local lattice variations that calculations may not account for. In this work we find surprisingly that a locally disordered material can exhibit a significantly larger Néel temperature (TN) than an ordered material of precisely the same chemical stoichiometry. Here, a EuFe2As2 crystal, which is a '122' parent of iron arsenide superconductors, is found through synthesis to have ordering below TN = 195 K (for the locally disordered crystal) or TN = 175 K (for the ordered crystal). In the higher TN crystals, there are shorter planar Fe-Fe bonds [2.7692(2) Å vs. 2.7745(3) Å], a randomized in-plane defect structure, and diffuse scattering along the [00 L] crystallographic direction that manifests as a rather broad specific heat peak. For the lower TN crystals, the a-lattice parameter is larger and the in-plane microscopic structure shows defect ordering along the antiphase boundaries, giving a larger TN and a higher superconducting temperature (Tc) upon the application of pressure. First-principles calculations find a strong interaction between c-axis strain and interlayer magnetic coupling, but little impact of planar strain on the magnetic order. Neutron single-crystal diffraction shows that the low-temperature magnetic phase transition due to localized Eu moments is not lattice or disorder sensitive, unlike the higher-temperature Fe sublattice ordering. This study demonstrates a higher magnetic ordering point arising from local disorder in 122.

| S-EPMC6934717 | biostudies-literature

Machine Learning Magnetic Parameters from Spin Configurations.

Project description:Hamiltonian parameters estimation is crucial in condensed matter physics, but is time- and cost-consuming. High-resolution images provide detailed information of underlying physics, but extracting Hamiltonian parameters from them is difficult due to the huge Hilbert space. Here, a protocol for Hamiltonian parameters estimation from images based on a machine learning (ML) architecture is provided. It consists in learning a mapping between spin configurations and Hamiltonian parameters from a small amount of simulated images, applying the trained ML model to a single unexplored experimental image to estimate its key parameters, and predicting the corresponding materials properties by a physical model. The efficiency of the approach is demonstrated by reproducing the same spin configuration as the experimental one and predicting the coercive field, the saturation field, and even the volume of the experiment specimen accurately. The proposed approach paves a way to achieve a stable and efficient parameters estimation.

| S-EPMC7435232 | biostudies-literature

Spin-lattice decoupling in a triangular-lattice quantum spin liquid.

Project description:A quantum spin liquid (QSL) is an exotic state of matter in condensed-matter systems, where the electron spins are strongly correlated, but conventional magnetic orders are suppressed down to zero temperature because of strong quantum fluctuations. One of the most prominent features of a QSL is the presence of fractionalized spin excitations, called spinons. Despite extensive studies, the nature of the spinons is still highly controversial. Here we report magnetocaloric-effect measurements on an organic spin-1/2 triangular-lattice antiferromagnet, showing that electron spins are decoupled from a lattice in a QSL state. The decoupling phenomena support the gapless nature of spin excitations. We further find that as a magnetic field is applied away from a quantum critical point, the number of spin states that interact with lattice vibrations is strongly reduced, leading to weak spin-lattice coupling. The results are compared with a model of a strongly correlated QSL near a quantum critical point.

| S-EPMC5904176 | biostudies-literature

Towards exact molecular dynamics simulations with machine-learned force fields.

Project description:Molecular dynamics (MD) simulations employing classical force fields constitute the cornerstone of contemporary atomistic modeling in chemistry, biology, and materials science. However, the predictive power of these simulations is only as good as the underlying interatomic potential. Classical potentials often fail to faithfully capture key quantum effects in molecules and materials. Here we enable the direct construction of flexible molecular force fields from high-level ab initio calculations by incorporating spatial and temporal physical symmetries into a gradient-domain machine learning (sGDML) model in an automatic data-driven way. The developed sGDML approach faithfully reproduces global force fields at quantum-chemical CCSD(T) level of accuracy and allows converged molecular dynamics simulations with fully quantized electrons and nuclei. We present MD simulations, for flexible molecules with up to a few dozen atoms and provide insights into the dynamical behavior of these molecules. Our approach provides the key missing ingredient for achieving spectroscopic accuracy in molecular simulations.

| S-EPMC6155327 | biostudies-literature

Imaging of nuclear magnetic resonance spin-lattice relaxation activation energy in cartilage.

Project description:Samples of human and bovine cartilage have been examined using magnetic resonance imaging to determine the proton nuclear magnetic resonance spin-lattice relaxation time, T1, as a function of depth within through the cartilage tissue. T1 was measured at five to seven temperatures between 8 and 38°C. From this, it is shown that the T1 relaxation time is well described by Arrhenius-type behaviour and the activation energy of the relaxation process is quantified. The activation energy within the cartilage is approximately 11 ± 2 kJ mol-1 with this notably being less than that for both pure water (16.6 ± 0.4 kJ mol-1) and the phosphate-buffered solution in which the cartilage was immersed (14.7 ± 1.0 kJ mol-1). It is shown that this activation energy increases as a function of depth in the cartilage. It is known that cartilage composition varies with depth, and hence, these results have been interpreted in terms of the structure within the cartilage tissue and the association of the water with the macromolecular constituents of the cartilage.

| S-EPMC6083713 | biostudies-literature

Machine-Learned Molecular Surface and Its Application to Implicit Solvent Simulations.

Project description:Implicit solvent models, such as Poisson-Boltzmann models, play important roles in computational studies of biomolecules. A vital step in almost all implicit solvent models is to determine the solvent-solute interface, and the solvent excluded surface (SES) is the most widely used interface definition in these models. However, classical algorithms used for computing SES are geometry-based, so that they are neither suitable for parallel implementations nor convenient for obtaining surface derivatives. To address the limitations, we explored a machine learning strategy to obtain a level set formulation for the SES. The training process was conducted in three steps, eventually leading to a model with over 95% agreement with the classical SES. Visualization of tested molecular surfaces shows that the machine-learned SES overlaps with the classical SES in almost all situations. Further analyses show that the machine-learned SES is incredibly stable in terms of rotational variation of tested molecules. Our timing analysis shows that the machine-learned SES is roughly 2.5 times as efficient as the classical SES routine implemented in Amber/PBSA on a tested central processing unit (CPU) platform. We expect further performance gain on massively parallel platforms such as graphics processing units (GPUs) given the ease in converting the machine-learned SES to a parallel procedure. We also implemented the machine-learned SES into the Amber/PBSA program to study its performance on reaction field energy calculation. The analysis shows that the two sets of reaction field energies are highly consistent with a 1% deviation on average. Given its level set formulation, we expect the machine-learned SES to be applied in molecular simulations that require either surface derivatives or high efficiency on parallel computing platforms.

| S-EPMC9132718 | biostudies-literature

OmicsDI is part of the ELIXIR infrastructure

OmicsDI is an Elixir interoperability service. Learn more ›

Tweets

OmicsDI Databases

PRIDE
PeptideAtlas
MassIVE
JPOST Repository
Physiome Model Repository

EGA
EVA
ENA
LINCS
PAXDB
Cell Collective

MetaboLights
Metabolomics Workbench
MetabolomeExpress
GNPS
BioModels
FAIRDOMHub

ArrayExpress
dbGaP
ExpressionAtlas
GEO
NODE

Information

Databases
Help
API
Contact us
Code on GitHub
Terms of use
Submit Data