CatalyticMLLM: A Graph-Text Multimodal Large Language Model for Catalytic Materials

TL;DR

This paper introduces QE-Catalytic-V2, a unified multimodal large language model that integrates property prediction and inverse design of catalytic materials, enabling more stable and effective closed-loop optimization workflows.

Contribution

The work presents a novel unified graph-text multimodal LLM for catalytic materials that combines property prediction and inverse design in a shared representation space.

Findings

Outperforms decoupled baselines in relaxed-energy prediction.

Effectively generates physically feasible CIF candidates conditioned on target properties.

Validates the unified model's superiority in inverse design and property prediction tasks.

Abstract

Property prediction and inverse structural design of catalytic materials are typically modeled as two independent tasks: the former predicts target properties from given structures, whereas the latter generates candidate structures according to desired properties. Although the decoupled paradigm facilitates the implementation of a ``generation--evaluation--screening'' workflow, the inconsistency between the generative model and the property prediction model in terms of representation spaces and training objectives can readily introduce data distribution shifts and evaluator bias, thereby limiting the stability of closed-loop optimization. In this work, we propose QE-Catalytic-V2, a unified graph--text multimodal large language model for catalytic materials, which integrates property prediction and inverse design within the same model and shared representation space. Under this unified framework, QE-Catalytic-V2 can not only perform reliable property prediction by leveraging three-dimensional structures and textual information, but also generate and screen physically feasible CIF candidates conditioned on target properties, thereby forming a closed-loop optimization workflow of ``inverse design--prediction--screening--redesign.'' Experimental results demonstrate that this unified paradigm outperforms decoupled baselines on both catalytic relaxed-energy prediction and inverse design tasks, validating the effectiveness of jointly modeling property prediction and structure generation within a single multimodal model.