Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Concept Videos

Heterogeneous Catalysis01:22

Heterogeneous Catalysis

Heterogeneous catalysis involves a catalyst in a different phase from the reactants. It is a process where the catalyst and the reactants are in distinct phases, typically solid and gas or liquid.Most heterogeneous catalysts are metals, metal oxides, or acids. The list includes transition metals like iron (Fe), cobalt (Co), nickel (Ni), palladium (Pd), platinum (Pt), chromium (Cr), manganese (Mn), tungsten (W), silver (Ag), and copper (Cu). These metals possess partially vacant d orbitals that...
Cooperative Allosteric Transitions01:58

Cooperative Allosteric Transitions

Cooperative allosteric transitions can occur in multimeric proteins, where each subunit of the protein has its own ligand-binding site. When a ligand binds to any of these subunits, it triggers a conformational change that affects the binding sites in the other subunits; this can change the affinity of the other sites for their respective ligands. The ability of the protein to change the shape of its binding site is attributed to the presence of a mix of flexible and stable segments in the...
Imperfections in Crystal Structure: Non-Stoichiometric Defects01:29

Imperfections in Crystal Structure: Non-Stoichiometric Defects

Non-stoichiometric defects refer to a type of defect in the crystal structure of a compound where the ratio of its constituent elements deviates from the ideal stoichiometric ratio. There are two main types of non-stoichiometric defects: metal excess defects and metal deficiency defects.Metal excess defects occur when there is a slight surplus of metal ions than what is required by the stoichiometric ratio of the compound. For example, heating a sodium chloride crystal in sodium vapor results...
Hybridization of Atomic Orbitals II03:35

Hybridization of Atomic Orbitals II

sp3d and sp3d 2 Hybridization

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

Deep Learning-Based Framework for Efficient Design of Multicomponent High Hardness High Entropy Alloys.

ACS applied materials & interfaces·2025
Same author

Detailed Reaction Kinetics for Hydrocarbon Fuels: The Development and Application of the ReaxFF<sub>CHO</sub>-S22 Force Field for C/H/O Systems with Enhanced Accuracy.

The journal of physical chemistry. A·2024
Same author

Carbon nanowires made by the insertion-and-fusion method toward carbon-hydrogen nanoelectronics.

Nanoscale·2023
Same author

High Hydrostatic Pressure Inducible Trimethylamine <i>N</i>-Oxide Reductase Improves the Pressure Tolerance of Piezosensitive Bacteria <i>Vibrio fluvialis</i>.

Frontiers in microbiology·2018
Same author

Protective role of melatonin in cardiac ischemia-reperfusion injury: From pathogenesis to targeted therapy.

Journal of pineal research·2018
Same author

Poly(Lactide-Co-Glycolide)-Monomethoxy-Poly-(Polyethylene Glycol) Nanoparticles Loaded with Melatonin Protect Adipose-Derived Stem Cells Transplanted in Infarcted Heart Tissue.

Stem cells (Dayton, Ohio)·2018

Related Experiment Video

Updated: May 14, 2026

Methods of Ex Situ and In Situ Investigations of Structural Transformations: The Case of Crystallization of Metallic Glasses
08:55

Methods of Ex Situ and In Situ Investigations of Structural Transformations: The Case of Crystallization of Metallic Glasses

Published on: June 7, 2018

Transfer Learning from Homogeneous to Heterogeneous: Fine-Tuning a Pretrained Interatomic Potential for

Lixin Fang1,2, Liqin Qin1,2, Limin Zhang1,2

  • 1Materials Genome Institute, Shanghai University, Shanghai 200444, China.

Materials (Basel, Switzerland)
|May 13, 2026
PubMed
Summary

This study fine-tuned a machine learning interatomic potential (MLIP) for molybdenum-based alloys, significantly improving accuracy for complex systems. Transfer learning shows promise but challenges remain for predicting entirely new elements.

Keywords:
MACE foundation modellocalized substitutional dopingmachine learning interatomic potentials (MLIPs)multicomponent Mo alloy designtransfer fine-tuning

More Related Videos

Multiscale Sampling of a Heterogeneous Water/Metal Catalyst Interface using Density Functional Theory and Force-Field Molecular Dynamics
10:52

Multiscale Sampling of a Heterogeneous Water/Metal Catalyst Interface using Density Functional Theory and Force-Field Molecular Dynamics

Published on: April 12, 2019

Discovery and Synthesis Optimization of Isoreticular Al(III) Phosphonate-Based Metal-Organic Framework Compounds Using High-Throughput Methods
07:20

Discovery and Synthesis Optimization of Isoreticular Al(III) Phosphonate-Based Metal-Organic Framework Compounds Using High-Throughput Methods

Published on: October 6, 2023

Related Experiment Videos

Last Updated: May 14, 2026

Methods of Ex Situ and In Situ Investigations of Structural Transformations: The Case of Crystallization of Metallic Glasses
08:55

Methods of Ex Situ and In Situ Investigations of Structural Transformations: The Case of Crystallization of Metallic Glasses

Published on: June 7, 2018

Multiscale Sampling of a Heterogeneous Water/Metal Catalyst Interface using Density Functional Theory and Force-Field Molecular Dynamics
10:52

Multiscale Sampling of a Heterogeneous Water/Metal Catalyst Interface using Density Functional Theory and Force-Field Molecular Dynamics

Published on: April 12, 2019

Discovery and Synthesis Optimization of Isoreticular Al(III) Phosphonate-Based Metal-Organic Framework Compounds Using High-Throughput Methods
07:20

Discovery and Synthesis Optimization of Isoreticular Al(III) Phosphonate-Based Metal-Organic Framework Compounds Using High-Throughput Methods

Published on: October 6, 2023

Area of Science:

  • Materials Science
  • Computational Materials Science
  • Machine Learning in Materials

Background:

  • Machine learning interatomic potentials (MLIPs) are typically trained on simple, ordered materials, limiting their use in complex alloys.
  • Existing MLIPs often fail for locally ordered heterogeneous systems (LOHetS), such as multicomponent alloys with substitutional elements.
  • Accurate modeling of alloy systems is crucial for designing new materials with desired properties.

Purpose of the Study:

  • To develop a fine-tuned MLIP capable of accurately describing doping in alloy systems, specifically Mo-based dilute alloys.
  • To evaluate the efficacy of transfer learning from globally ordered homogeneous systems (GOHomS) to LOHetS.
  • To assess the performance of the fine-tuned MLIP in predicting energies and forces for multicomponent alloys.

Main Methods:

  • Developed a fine-tuned MLIP using the MACE foundation model for Mo-based alloys with up to 20 substitutional elements.
  • Trained the model on over 7000 structures generated using first-principles density functional theory (DFT) calculations.
  • Evaluated model accuracy using mean absolute error (MAE) and root-mean-square error (RMSE) for energy and force predictions.

Main Results:

  • The fine-tuned MLIP achieved state-of-the-art accuracy with MAE of 2.27 meV/atom and RMSE of 3.79 meV/atom for energy.
  • Force predictions showed MAE of 13.83 meV/Å and RMSE of 24.26 meV/Å.
  • Fine-tuned models demonstrated 7-20 times improvement over models trained from scratch or zero-shot foundation models, with challenges in extrapolating to unknown elements.

Conclusions:

  • Transfer learning is effective for applying MLIPs from GOHomS to LOHetS in multicomponent alloys.
  • Including unknown elements in training datasets is crucial for high accuracy without additional computational cost.
  • Further data and fine-tuning may be required for more complex alloy systems beyond dilute Mo-based alloys.