ICN2 Joins the LLM Hackathon 2025 with a New Tool for Nanoparticle Analysis

by Chema Arcos

A collaborative project between the ICN2 Artificial Intelligence Platform and the Inorganic Nanoparticles Group has resulted in the development of an innovative approach that combines electron microscopy, computer vision, and artificial intelligence.

Last week, ICN2 took part in the 2025 edition of the LLM Hackathon for Applications in Materials Science & Chemistry. This international event, which focused on using artificial intelligence to accelerate discoveries and develop new strategies in chemistry and materials science, was held online across different parts of the world.

The ICN2 team comprised the AI Platform (Andrés Henao and Francisco A. Molina) and the Inorganic Nanoparticles Group (Dr Neus Bastús and Dr Muriel Freixanet). They presented ms2nano:nanoparticle-image-analysis, an open-source project for analysing transmission electron microscopy (TEM) images. This tool enables morphological features of nanoparticles to be extracted and correlated with catalytic activity in chemical reactions.

During the hackathon, the team also introduced a new functionality based on large language models (LLMs), a type of AI designed to work with vast amounts of information that helps to simplify and improve the interpretation of results.

Supported by the Severo Ochoa Programme and the ICN2 Seed Funding Programme, this project highlights the institute’s strong commitment to Open Science.

Disclaimer: The image was generated using AI (Gemini 2.5 Flash model).