The rapid advancement of artificial intelligence has transformed materials science research, with large language models (LLMs) emerging as a pivotal driver of innovation. This review explores the comprehensive role of LLMs in accelerating materials design across the entire research lifecycle, from knowledge mining to intelligent design. The study aims to highlight how LLMs can address challenges in traditional materials research, such as data fragmentation, high experimental costs, and limited reasoning capabilities, by leveraging their strengths in information retrieval, cross-modal data integration, and intelligent reasoning.
Key methodologies include the application of LLMs in knowledge discovery through techniques like retrieval-augmented generation (RAG), multi-modal information retrieval, and knowledge graph construction. These approaches enable efficient extraction and structuring of materials data from vast repositories of scientific literature and experimental records. Additionally, LLMs are integrated with automated experimental platforms to optimize workflows, from natural language-driven experiment design to high-throughput iterative testing.
The results demonstrate that LLMs significantly enhance materials research efficiency and accuracy. For instance, in knowledge mining, LLMs improve information retrieval precision by up to 29.4% in tasks like predicting material synthesis conditions. In materials design, LLMs enable accelerated computational modeling, structural and property prediction, and inverse design, reducing experimental trial-and-error cycles. Notably, LLMs excel in cross-scale knowledge integration, linking material composition, processing parameters, and performance metrics to guide innovative synthesis pathways.
However, challenges persist, including the reliance on high-quality data, the "black-box" nature of LLMs, and limitations in handling complex material systems. Future directions emphasize enhancing data quality through multi-source integration, improving model explainability via visualization tools, and deepening interdisciplinary collaboration to bridge gaps between AI and domain-specific expertise.
In conclusion, LLMs are reshaping materials science by enabling data-driven, knowledge-intensive research paradigms. Their ability to integrate vast datasets, predict material properties, and automate experimental workflows positions them as indispensable tools for accelerating materials discovery and innovation. As LLMs evolve, their synergy with physical constraints and experimental platforms promises to unlock new frontiers in materials design.