In the realm of energy research, a team of scientists from the Department of Energy’s Lawrence Berkeley National Laboratory has developed a novel tool aimed at streamlining computational heterogeneous catalysis studies. The researchers, Honghao Chen, Jiangjie Qiu, Yi Shen Tew, and Xiaonan Wang, have introduced CatMaster, an agentic autonomous system designed to simplify and accelerate the complex workflows involved in catalytic research.
Computational heterogeneous catalysis often relies on density functional theory (DFT) to link atomic structures with catalytic behavior. However, these studies can be hindered by lengthy, costly, and iterative processes that are sensitive to setup choices. Practical issues such as maintaining consistent references, preparing multiple related inputs, recovering from failed runs, and keeping a complete record of actions can slow down projects and make results difficult to reproduce or extend.
CatMaster addresses these challenges by transforming natural language requests into complete calculation workspaces. This includes structures, inputs, outputs, logs, and a concise run record. The system maintains a persistent project record of key facts, constraints, and file pointers to support inspection and restartability. It is paired with a multi-fidelity tool library that covers rapid surrogate relaxations and high-fidelity DFT calculations for validation when needed.
The researchers demonstrated CatMaster on four increasingly complex tasks: an O2 spin-state check with remote execution, BCC Fe surface energies with a protocol-sensitivity study and CO adsorption site ranking, high-throughput Pt–Ni–Cu alloy screening for hydrogen evolution reaction (HER) descriptors with surrogate-to-DFT validation, and a demonstration beyond the predefined tool set, including equation-of-state fitting for BCC Fe and CO-FeN4-graphene single-atom catalyst geometry preparation.
By reducing manual scripting and bookkeeping while keeping the full evidence trail, CatMaster aims to help catalysis researchers focus on modeling choices and chemical interpretation rather than workflow management. This tool has the potential to significantly enhance the efficiency and reproducibility of catalytic research, ultimately accelerating the development of new and improved catalysts for various energy applications.
The research was published in the journal Nature Computational Science.
This article is based on research available at arXiv.