Reshaping MOFs Text Mining with a Dynamic Multi-Agent Framework of Large Language Agents

2025-04-26

AI Summary

Precise extraction of metal–organic framework (MOF) synthesis conditions from vast scientific literature remains challenging due to heterogeneity, ambiguity, and lack of structured representation. Method: We propose MOFh6, a dynamic multi-agent framework powered by GPT-4o-mini that integrates multimodal inputs—including literature text, MOF structural encodings, and physicochemical properties—and orchestrates domain-specialized agents for synthesis planning, property prediction, and chemical reasoning. The system leverages a curated MOF knowledge graph and advanced prompt engineering to enable natural-language-driven, end-to-end synthesis pathway recommendation and automated generation of DFT-ready structural files (e.g., CIF). Contribution/Results: MOFh6 introduces the first dynamic multi-agent architecture for MOF synthesis informatics, significantly improving retrieval accuracy and efficiency of synthesis conditions. It supports heterogeneous query formats and achieves full automation—from unstructured textual descriptions to computationally executable models—marking the first such capability in MOF computational design.

Technology Category

• Machine Learning: Mixture of Experts (MoE)
• Multiagent Systems: Multiagent Planning
• Cognitive Modeling & Cognitive Systems: Agent Architectures

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Search and Retrieval-Augmented AI: Agentic search
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs

Abstract

The mining of synthesis conditions for metal-organic frameworks (MOFs) is a significant focus in materials science. However, identifying the precise synthesis conditions for specific MOFs within the vast array of possibilities presents a considerable challenge. Large Language Models (LLMs) offer a promising solution to this problem. We leveraged the capabilities of LLMs, specifically gpt-4o-mini, as core agents to integrate various MOF-related agents, including synthesis, attribute, and chemical information agents. This integration culminated in the development of MOFh6, an LLM tool designed to streamline the MOF synthesis process. MOFh6 allows users to query in multiple formats, such as submitting scientific literature, or inquiring about specific MOF codes or structural properties. The tool analyzes these queries to provide optimal synthesis conditions and generates model files for density functional theory pre modeling. We believe MOFh6 will enhance efficiency in the MOF synthesis of all researchers.

Problem

• Identifying precise MOF synthesis conditions from vast possibilities
• Integrating MOF-related agents using LLMs for synthesis optimization
• Streamlining MOF synthesis queries and DFT pre-modeling file generation

Innovation

• Dynamic multi-agent framework with LLMs
• Integration of MOF-related specialized agents
• Automated synthesis condition analysis and modeling