Mathematical & physical sciences · OECD FoS 2.5 · China MoE 0805 材料科学与工程
Materials science
材料科学
L2Local breakthroughs → Steady Score 2.20 · Last reviewed 2026-10-21 · Next review 2027-01-19
BIGGEST BREAKTHROUGH 2023-11-29 AI leapField · major
GNoME predicts 2.2M new crystals, ~380k stable (Nature) ↗
2.2M crystals, ~380k stable; novelty later questioned
Why this level
GNoME predicted 2.2M crystals and MatterGen designs materials generatively (Nature 2023/2025), A-Lab synthesises autonomously; but the A-Lab paper was corrected in Jan 2026 (41/58 → 36/57), so “AI-discovered materials” claims remain contested.
S = 0.30·method + 0.35·discovery + 0.25·adoption + 0.10·education (education scores 0 when there is no evidence; if empty, the other weights are renormalised)
Sources
- nature.com/articles/s41586-023-06735-9
- cen.acs.org/research-integrity/Nature-robot-chemist-paper-corrected/104/web/2026/01
Composite 1.5–2.49: at least one peer-reviewed/officially verified AI-key result, or one dimension at 3; routine workflow largely unchanged. All levels →
Research directions
Crystal structure predictionGenerative materials designAutonomous synthesisML interatomic potentials
Papers
- AI-accelerated scalable synthesis of nanocrystals for cost-effective photovoltaics
AI: ML plus automated flow synthesis optimise the process; Field: pushes lab materials toward low-cost scale-up
2026-10-02 · Nature Communications
- Bayesian
Conavigation of a Computational Physical
Model and Atomic Force Microscopy Experiment to Autonomously Survey
a Combinatorial Materials Library
Method improvement
2026-08-17 · ACS Nano
- Machine learning driven discovery of low modulus biomedical titanium alloys for additive manufacturing
AI-designed alloy validated for manufacturable implants
2026-04-14 · Nature Communications
- DiffSyn: a generative diffusion approach to materials synthesis planning
From predicting structures to planning how to make them
2026-02-02 · Nature Computational Science
- Developing a complete AI-accelerated workflow for superconductor discovery
AI-predicted superconductors confirmed in the lab
2026-01-27 · npj Computational Materials
- A generative model for inorganic materials design
Generative design with one synthesised validation
2025-01-16 · Nature
- A foundation model for atomistic materials chemistry
One model simulates most inorganic systems
2023-12-29 · arXiv
-
AI leap
Scaling deep learning for materials discovery2.2M crystals, ~380k stable; novelty later questioned
2023-11-29 · Nature
- An autonomous laboratory for the accelerated synthesis of inorganic materials
Autonomous synthesis; corrected to 36/57 in Jan 2026
2023-11-29 · Nature
- Compositional neural networks for multimodal X-ray and neutron scattering inversion
Compositional neural networks combine X-ray and neutron scattering to reduce non-uniqueness in structural inversion.
2026-10-21 · arXiv · Valentin Munteanu, Alexander Hinderhofer, Dmitry Lapkin, Vladimir Starostin
- GBClaw: An agentic AI framework for atomistic grain boundary research and mechanistic discovery
An LLM agent framework automating atomistic studies of general grain boundaries; applied to 14 Ni boundaries to identify deformation modes.
2026-10-20 · arXiv · Yazhuo Liu, Kunqing Ding, Xiaoguo Gong, Zifeng Wang
- Symmetry-Guided Machine Learning of Spin-Dependent Nonlinear Hall Response
A symmetry-constrained model predicts the Berry-curvature dipole for held-out spin configurations with R² > 0.9995.
2026-10-20 · arXiv · Jyotirmoy Sau
- TexFlow: learning crystallographic texture on SO(3) manifold via normalizing flows
A conditional normalising flow on SO(3) that models crystallographic texture for process–structure–property prediction.
2026-10-20 · arXiv · Gowtham Nimmal Haribabu, Tarakram Ramgopal, Divakar Vetriveeran, Siddhant Kumar
- CrystalAR: Autoregressive Modeling for Symmetric Crystal Generation
An autoregressive model jointly generating Wyckoff templates and full crystal structures via hierarchical discrete localisation.
2026-10-20 · arXiv · Ziqiao Meng, Rongsheng Chen, Zhi Zheng, Nikita Kazeev
- MOFFieldBench: A Machine Learning Benchmark for Continuous Electronic Fields in Metal-Organic Frameworks
A benchmark of electron-density and electrostatic-potential fields for learning representations of metal–organic frameworks.
2026-10-19 · arXiv · Zhaolin Hu, Hehe Fan, Chenhao Rao, Wangyihan Guo
- Machine Learning-Assisted Discovery of Carbon-Related Defect Formation Mechanisms in Diamond
ML-assisted simulations reveal formation mechanisms of di-interstitials and di-vacancies in diamond, relevant to spin qubits.
2026-10-19 · arXiv · Grace Huang, Christopher Fivecoat, Xun Xu, Mengen Wang
- Crystal Structure Based Graph Neural Network And LLM Multi-Modal Approach for Material Magnetic Ordering Prediction
Combines crystal-graph neural networks and an LLM to predict magnetic ordering from crystal structure.
2026-10-19 · arXiv · Adam Zheng, April Zheng, Sheng Ran
- From arXiv, bioRxiv, medRxiv and OpenAlex (last ~30 days); summaries written by AIMap from the abstract.
- OR-License: Evaluating LLM Agents as Real-World Operations Research Practitioners 2026-10-16
- BlenderWorld: Benchmarking Visual Agents for 3D Creation from Real-World References 2026-10-14
- Into the Unknown: Benchmarking AI agents on open scientific research problems 2026-10-13
- All related papers in the archive →
Milestones 10· 1 leaps
Big glowing nodes are leaps: AI leap cool · Field leap warm · double ring = both
-
Paper
AI-accelerated scalable synthesis of nanocrystals for cost-effective photovoltaics (Nat. Commun.)AI: ML plus automated flow synthesis optimise the process; Field: pushes lab materials toward low-cost scale-up
Nature Communications -
Paper
Bayesian conavigation couples AFM experiments with a kMC model (ACS Nano)Method improvement
ACS Nano -
Paper
ML-designed AM-specific low-modulus Ti alloy validated by laser powder-bed fusionAI-designed alloy validated for manufacturable implants
Nature Communications -
Paper
DiffSyn plans synthesis routes; a new zeolite with better thermal stability is madeFrom predicting structures to planning how to make them
Nature Computational Science -
Paper
A-Lab paper corrected: successes revised to 36 of 57Correction shows need for stricter validation
C&EN -
Paper
BEE-NET screens 1.3M structures; two previously unknown superconductors confirmed experimentallyAI-predicted superconductors confirmed in the lab
npj Computational Materials -
Paper
MatterGen generative design of inorganic materials (Nature)Generative design with one synthesised validation
Nature -
System
MACE-MP-0: a foundation ML interatomic potential across the periodic tableOne model simulates most inorganic systems
arXivPreprint -
PaperAI leap
GNoME predicts 2.2M new crystals, ~380k stable (Nature)2.2M crystals, ~380k stable; novelty later questioned
Nature -
Paper
A-Lab synthesises new compounds in 17 days (Nature; corrected 2026)Autonomous synthesis; corrected to 36/57 in Jan 2026
NatureContested
No milestones match this filter.
More papers & repos 10
| Type | Title | Date |
|---|---|---|
| Paper | AI-accelerated scalable synthesis of nanocrystals for cost-effective photovoltaics | 2026-10-02 |
| Paper | Machine learning driven discovery of low modulus biomedical titanium alloys for additive manufacturing | 2026-04-14 |
| Paper | DiffSyn: a generative diffusion approach to materials synthesis planning | 2026-02-02 |
| Paper | Developing a complete AI-accelerated workflow for superconductor discovery | 2026-01-27 |
| Paper | A generative model for inorganic materials design | 2025-01-16 |
| Paper | A foundation model for atomistic materials chemistry Preprint | 2023-12-29 |
| Paper | Scaling deep learning for materials discovery | 2023-11-29 |
| Paper | An autonomous laboratory for the accelerated synthesis of inorganic materials | 2023-11-29 |
| Repo | google-deepmind/materials_discovery (GNoME) To be verified | — |
| Repo | microsoft/mattergen To be verified | — |
Latest news 8
- Multi-property Feynman-Kac Steering of Crystal Diffusion ModelsarXiv cond-mat.mtrl-scienPaperPreprint
- Crystal Structure Based Graph Neural Network And LLM Multi-Modal Approach for Material Magnetic Ordering PredictionarXiv cond-mat.mtrl-scienPaperPreprint
- Data-Efficient Indentation Size Effect Mitigation in Steels Using Machine Learning and a Physics-Constrained Neural NetworkarXiv cond-mat.mtrl-scienPaperPreprint
- Ken Crutchfield: Lessons From Steel — Why LLMs Will Become CommoditiesLawNext / LawSitesen