Hierarchical Multi-agent Large Language Model Reasoning for Autonomous Heterogeneous Catalyst Discovery
- 1School of Civil & Environmental Engineering, University of Connecticut, Storrs, CT, USA
- 2Theoretical Division, Los Alamos National Laboratory, Los Alamos, NM, USA
Abstract
Artificial intelligence is reshaping scientific exploration, but most methods automate procedural tasks without engaging in scientific reasoning, limiting autonomy in discovery. We demonstrate that hierarchical agentic large language model reasoning can efficiently drive simulation and scientific exploration. Across two chemical applications, CO adsorption on Cu surface transition metal adatoms and on M–N–C catalysts, reasoning-guided exploration reduces required atomistic simulations by up to 90% relative to heuristic or random selection. Comparisons across single-agent, multi-agent, and stochastic baselines show that hierarchical strategies yield more coherent and information-efficient search trajectories. Reasoning traces reveal chemically grounded decisions that cannot be explained by semantic bias or stochastic sampling. We realize these agentic reasoning strategies in Materials Agents for Simulation and Theory in Electronic-structure Reasoning (MASTER), a multimodal system that translates natural language into density functional theory workflows. Altogether, multi-agent collaboration accelerates heterogeneous catalyst discovery and marks a step toward more autonomous, reasoning-guided scientific exploration.
Related
- Autonomous in-silico inorganic materials discovery via multi-agent physics-aware scientific reasoning
- Constructive chemical discovery through autonomous reasoning and experiment
- Agentic AI for Scientific Reasoning in Autonomous Quantum Sensing Experiments
- Surfactant-Toolkit and SurfBench-200: an open computation toolkit and a benchmark of tool-augmented language-model reasoning in surfactant science
- How Chemical Evidence Shapes Mechanistic Reasoning and Prediction by Large Language Models: Fine-Grained Inference in Nucleoside-Derived Supramolecular Gelation
BibTeX
@article{rothfarb2026hierarchical,
title = {{Hierarchical Multi-agent Large Language Model Reasoning for Autonomous Heterogeneous Catalyst Discovery}},
author = {Samuel Rothfarb and Megan Christina Davis and Ivana Matanović and Baikun Li and Edward F. Holby and Wilton J. M. Kort-Kamp},
journal = {npj Computational Materials},
year = {2026},
month = may,
doi = {10.1038/s41524-026-02139-1},
url = {https://doi.org/10.1038/s41524-026-02139-1},
note = {CC BY-NC-ND},
}
RIS
TY - JOUR
TI - Hierarchical Multi-agent Large Language Model Reasoning for Autonomous Heterogeneous Catalyst Discovery
AU - Samuel Rothfarb
AU - Megan Christina Davis
AU - Ivana Matanović
AU - Baikun Li
AU - Edward F. Holby
AU - Wilton J. M. Kort-Kamp
PY - 2026
DA - 2026/05/25
JO - npj Computational Materials
DO - 10.1038/s41524-026-02139-1
UR - https://doi.org/10.1038/s41524-026-02139-1
AB - Artificial intelligence is reshaping scientific exploration, but most methods automate procedural tasks without engaging in scientific reasoning, limiting autonomy in discovery. We demonstrate that hierarchical agentic large language model reasoning can efficiently drive simulation and scientific exploration. Across two chemical applications, CO adsorption on Cu surface transition metal adatoms and on M–N–C catalysts, reasoning-guided exploration reduces required atomistic simulations by up to 90% relative to heuristic or random selection. Comparisons across single-agent, multi-agent, and stochastic baselines show that hierarchical strategies yield more coherent and information-efficient search trajectories. Reasoning traces reveal chemically grounded decisions that cannot be explained by semantic bias or stochastic sampling. We realize these agentic reasoning strategies in Materials Agents for Simulation and Theory in Electronic-structure Reasoning (MASTER), a multimodal system that translates natural language into density functional theory workflows. Altogether, multi-agent collaboration accelerates heterogeneous catalyst discovery and marks a step toward more autonomous, reasoning-guided scientific exploration.
SN - 2057-3960
ER -