LeMat-Synth: a multi-modal toolbox to curate broad synthesis procedure databases from scientific literature

TL;DR

LeMat-Synth is a comprehensive multi-modal toolbox that uses advanced AI models to extract, organize, and structure synthesis procedures from a vast collection of scientific literature, facilitating materials discovery.

Contribution

It introduces a novel AI-driven approach to systematically extract and organize synthesis data from literature, creating a large, structured dataset and open-source tools for the community.

Findings

Curated 81,000 papers into a structured dataset of synthesis procedures.

Achieved high extraction quality validated by expert annotations.

Provided an open-source software library for community extension.

Abstract

The development of synthesis procedures remains a fundamental challenge in materials discovery, with procedural knowledge scattered across decades of scientific literature in unstructured formats that are challenging for systematic analysis. In this paper, we propose a multi-modal toolbox that employs large language models (LLMs) and vision language models (VLMs) to automatically extract and organize synthesis procedures and performance data from materials science publications, covering text and figures. We curated 81k open-access papers, yielding LeMat-Synth (v 1.0): a dataset containing synthesis procedures spanning 35 synthesis methods and 16 material classes, structured according to an ontology specific to materials science. The extraction quality is rigorously evaluated on a subset of 2.5k synthesis procedures through a combination of expert annotations and a scalable LLM-as-a-judge framework. Beyond the dataset, we release a modular, open-source software library designed to support community-driven extension to new corpora and synthesis domains. Altogether, this work provides an extensible infrastructure to transform unstructured literature into machine-readable information. This lays the groundwork for predictive modeling of synthesis procedures as well as modeling synthesis--structure--property relationships.