LLM-Driven Knowledge Extraction for Noble Metal Nanomaterial Discovery and Applications

Advances in the discovery, synthesis, characterisation and deployment of noble metal nanomaterials, including gold, silver, platinum and palladium, are central to transformative progress across a wide range of sectors, notably catalysis, biomedicine, chemical sensing and energy conversion and storage.

Despite rapid growth in the field, the pace of innovation is increasingly limited by the fragmentation of knowledge across disparate sources, including peer-reviewed literature, experimental datasets, patents and computational materials databases. Valuable insights are often buried within unstructured text or siloed resources, hindering systematic comparison, reuse, and translation across disciplines.

This presentation discusses the development of a masked-language-model-based data retrieval and analysis pipeline capable of automatically extracting, structuring and synthesising information from the global corpus of noble metal nanomaterials research. By leveraging recent advances in artificial intelligence, natural language understanding and data-driven materials science, the approach aims to accelerate materials discovery by identifying underexplored compositions, morphologies, structure–property relationships and emerging application spaces. Ultimately, this work will provide a scalable and extensible foundation for targeted experimental validation and cross-domain innovation within the advanced material ecosystem, supporting both fundamental scientific discovery and applications of global relevance.