Mathematical & physical sciences · OECD FoS 1.4 · China MoE 0703 化学
Chemistry
化学
L3Systemic transformation ↑ Rising ↑ from L2 Score 2.55 · Last reviewed 2026-10-21 · Next review 2026-12-20
BIGGEST BREAKTHROUGH 2024-10-09 AI leapField leap
Chemistry Nobel to Baker, Hassabis & Jumper ↗
First chemistry Nobel for AI-driven work
Why this level
Raised to L3: the 2024 chemistry Nobel went to AI structure prediction and design; retrosynthesis and ML potentials are routine; in 2026 autonomous platforms discovered new reactions and ligands (Nat. Commun. Jun & Oct 2026) and low-cost self-driving labs appeared. Day-to-day lab work is still largely manual.
↑ L2 → L3
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
- nobelprize.org/prizes/chemistry/2024/summary
- nature.com/articles/s41467-026-78271-9
- nature.com/articles/s41467-026-74425-x
Composite 2.5–3.49 (or ≥3.5 with any of method/discovery/adoption below 3): AI is standard infrastructure with multiple independent breakthroughs. All levels →
Research directions
Self-driving labsReaction prediction & retrosynthesisChemistry LLM agents
Papers
- Reinforcement learning guided robotic platform that investigates and discovers asymmetric reactions
Autonomous discovery of new asymmetric reactions, not just optimisation
2026-10-13 · Nature Communications
- An autonomous lab for data-driven homogeneous catalysis
Moves from tuning to discovering new catalyst behaviour, validated at 10× scale
2026-06-20 · Nature Communications
- Autonomous microfluidic experimentation for exploring reaction inference and synthesizing double perovskite nanoplatelets
Microfluidic SDL optimises six-element nanoplatelets
2026-05-04 · Nature Communications
- A flexible and affordable self-driving laboratory for automated reaction optimization
Puts closed-loop optimisation within reach of ordinary labs
2026-04-13 · Nature Synthesis
- Augmenting large language models with chemistry tools
LLM agent executes real syntheses
2024-05-08 · Nature Machine Intelligence
- Autonomous chemical research with large language models
First LLM system to autonomously plan and run real chemistry experiments
2023-12-20 · Nature
- A mobile robotic chemist
First mobile autonomous robot chemist in a standard lab
2020-07-08 · Nature
- Planning chemical syntheses with deep neural networks and symbolic AI
In blind tests chemists could not tell AI routes from literature routes
2018-03-29 · Nature
- Breaking Bottlenecks in Solid Electrolyte Discovery With Large Artificial Intelligence Models
Outlines a framework for autonomous solid-electrolyte discovery driven by large AI models across transport, defect, mechanical and interface objectives.
2026-10-20 · OpenAlex · KITopen · Eric Jianfeng Cheng, Min Hong, Zhiquan Zeng, Chuanyu Liu
- Magnolia: A Persistent-Memory Agent Framework for Computational Chemistry
A persistent-memory LLM agent for computational chemistry that distils observations into human-validated cross-project rules.
2026-10-19 · arXiv · Tao Jiang, Benjamin Navet, Yousouf Taghzouti, Fabien Gandon
- Challenging Atomistic Foundation Models on Disordered Carbons
Stress-tests six popular foundation ML interatomic potentials on disordered carbons to assess out-of-domain reliability.
2026-10-19 · arXiv · Joseph Hart, Samuel G. H. Brookes, Fabian L. Thiemann, Christoph Schran
- Global Research Trends in Artificial Intelligence and Machine Learning for Drug Delivery Technology: A Bibliometric and Science Mapping Analysis (2015–2026)
Bibliometric mapping of AI/ML research in drug delivery (2015–2026).
2026-10-18 · OpenAlex · International Journal of Drug Delivery Technology
- Aptamers as a sandbox for lab-in-the-loop agentic molecular discovery
Uses aptamer design (1–2 week cycles, $1–3 per candidate) to test whether LLM agents iterating with wet-lab data speed up de novo discovery.
2026-10-17 · arXiv · Zachary Cutts, John Coller, Rachel Chen, Hua Tang
- Experimental Design Signatures Emerge from Topic Modeling of Graphene Oxide Literature
Topic modelling of the vast graphene-oxide literature reveals experimental design signatures.
2026-10-17 · OpenAlex · ACS Nano · Henry Julio Kobs, Henrique Ferreira, Yang Liu, Amauri Jardim de Paula
- SpecLLM: LLM-based VNIR spectral interpretation
Makes VNIR spectra accessible to LLMs through feature-level descriptions for interpretation.
2026-10-17 · OpenAlex · Scientific Reports · Michał Romaszewski, Anna Strzoda, Marzena Halama, Przemysław Głomb
- From arXiv, bioRxiv, medRxiv and OpenAlex (last ~30 days); summaries written by AIMap from the abstract.
Milestones 9· 1 leaps
Big glowing nodes are leaps: AI leap cool · Field leap warm · double ring = both
-
Paper
RL-guided robot makes 3,000+ decisions in 30 days and discovers four highly enantioselective reactions (Nat. Commun.)Autonomous discovery of new asymmetric reactions, not just optimisation
Nature Communications -
Paper
Flex-Cat autonomous lab: 680 experiments, programmable “flexible ligands”, >2.5× turnover (Nat. Commun.)Moves from tuning to discovering new catalyst behaviour, validated at 10× scale
Nature Communications -
Paper
PoLARIS self-driving microfluidic lab optimises double-perovskite nanoplatelets (Nat. Commun.)Microfluidic SDL optimises six-element nanoplatelets
Nature Communications -
System
RoboChem-Flex: a ~US$5,000 open self-driving chemistry lab (Nature Synthesis)Puts closed-loop optimisation within reach of ordinary labs
Nature Synthesis -
PrizeAI leapField leap
Chemistry Nobel to Baker, Hassabis & JumperFirst chemistry Nobel for AI-driven work
Nobel Prize -
Paper
ChemCrow augments an LLM with 18 chemistry tools (Nat. Mach. Intell.)LLM agent executes real syntheses
Nature Machine Intelligence -
Paper
Coscientist: LLM autonomously plans and runs chemistry experiments (Nature)First LLM system to autonomously plan and run real chemistry experiments
Nature -
Paper
A mobile robotic chemist runs 688 experiments in 8 days and finds better photocatalysts (Nature)First mobile autonomous robot chemist in a standard lab
Nature -
Paper
Planning chemical syntheses with deep neural networks and symbolic AI (Nature)In blind tests chemists could not tell AI routes from literature routes
Nature
No milestones match this filter.
More papers & repos 9
| Type | Title | Date |
|---|---|---|
| Paper | Reinforcement learning guided robotic platform that investigates and discovers asymmetric reactions | 2026-10-13 |
| Paper | An autonomous lab for data-driven homogeneous catalysis | 2026-06-20 |
| Paper | Autonomous microfluidic experimentation for exploring reaction inference and synthesizing double perovskite nanoplatelets | 2026-05-04 |
| Paper | A flexible and affordable self-driving laboratory for automated reaction optimization | 2026-04-13 |
| Paper | Augmenting large language models with chemistry tools | 2024-05-08 |
| Paper | ChemCrow: Augmenting LLMs with chemistry tools (Nat. Mach. Intell. 2024) To be verified | 2024-05 |
| Paper | Autonomous chemical research with large language models | 2023-12-20 |
| Paper | A mobile robotic chemist | 2020-07-08 |
| Paper | Planning chemical syntheses with deep neural networks and symbolic AI | 2018-03-29 |
Latest news 12
- Aptamers as a sandbox for lab-in-the-loop agentic molecular discoveryarXiv q-bio.BMenPaperPreprint
- Distillation, Self-Training, and the Fragility of Learned Chemical ReasoningarXiv cs.AIenPaperPreprint
- Measurement-efficient energy estimator for quantum chemistry of photodynamic materialsarXiv physics.chem-phenPaperPreprint
- Designing enzymes for new-to-nature chemistry and non-natural substrates with AlphaProtein NovobioRxiv (synthetic biology)enPreprint